representation without parties: reconsidering the one-party...
TRANSCRIPT
Representation without Parties:Reconsidering the One-Party South
Devin Caughey∗
November 19, 2014
Abstract
The one-party South is considered the canonical example of democracy’s im-possibility in the absence of partisan competition. I argue that this viewconflates the effects of disfranchisement and one-partyism. Despite the lack ofpartisan competition, Democratic primaries induced a “selectoral connection”between Southern members of Congress and the (white) eligible electorate. Us-ing survey and roll-call data from the 1930s–50s, I show that Southern whiteswere collectively well represented in Congress. Southern MCs’ transformationfrom New Dealers to centrists mirrored analogous shifts in the white public.Cross-sectionally, Southern senators were just as responsive to their electorates’economic policy preferences as non-Southern senators. Data on income differ-ences across electorates indicate that Southern representatives were responsivetoo, though less so than senators. In addition to illuminating a pivotal periodin American political, these findings suggest that partisan competition is notnecessary for effective representation.
∗Political Science Department, Massachusetts Institute of Technology ([email protected]). Forfeedback and assistance, I thank Eric Schickler, Chris Warshaw, Adam Berinsky, Melissa Meek,Ira Katznelson, Rob Mickey, David Mayhew, Pat Egan, Jas Sekhon, Laura Stoker, Kevin Quinn,Sean Farhang, Jim Alt, Mark Franklin, Thad Kousser, Mona Vakilifathi, David Wiens, and seminarparticipants at UC Berkeley, MIT, Yale, University of Illinois, Princeton, University of Chicago,Cornell, UCSD, and APSA. I am also grateful to Jim Snyder, Steven Ansolabehere, Mark Hansen,and Shigeo Hirano for sharing their data on Southern primary elections.
1
Contents
1 Introduction 3
2 The One-Party South, 1930s–60s 52.1 Southern Democrats in Congress . . . . . . . . . . . . . . . . . . . . 7
3 The Selectoral Connection in the One-Party South 103.1 Participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.2 Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.3 Responsiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4 Statistical Evidence for Representation 164.1 New Deal Liberalism in the Public . . . . . . . . . . . . . . . . . . . 164.2 Collective Representation . . . . . . . . . . . . . . . . . . . . . . . . . 194.3 Responsiveness to Economic Policy Preferences . . . . . . . . . . . . 204.4 Responsiveness to Economic Interests . . . . . . . . . . . . . . . . . . 27
5 Conclusion 32
A Estimation and Imputation of Income 45
B Supplementary Figures 47
2
1 Introduction
Democracy is widely considered to require partisan competition. Among politicalscientists, support for E. E. Schattschneider’s famous claim that that “democracyis unthinkable” without parties is as close to unanimous as the discipline gets.1
Some works, such as Alvarez et al. (1996), go so far as to incorporate partisancompetition into their very definition of democracy. But even scholars who considerthe two concepts logically separable believe that, empirically, “partisan competitionis a necessary precondition for democracy” (Aldrich and Griffin 2010, 595; see alsoAmerican Political Science Association 1950; Lipset 1996; Katz 2006; Aldrich 2011).
Most arguments for democracy’s dependence on party competition hinge on par-ties’ role in providing information to citizens.2 In this view, the structure and con-tinuity provided by party labels enable voters to make informed choices betweencompeting candidates and to hold incumbents accountable for their behavior in of-fice (Downs 1957; Snyder and Ting 2002; Wright and Schaffner 2002; Aldrich andGriffin 2010). Without these aids to prospective choice and retrospective account-ability, citizens cannot induce governments to be responsive to their preferences.3
Since representation is a key characteristic of and justification for democracy (Mill1867; Dahl 1971; Sen 1999), lack of partisan competition thus is fatal to democracy,or at least to a well-functioning one.4
My goal here is to dispute the notion that electoral competition between party-affiliated candidates is necessary for representation. I further argue that under cer-tain institutional conditions, nonpartisan elections are just as effective at promoting
1. “No theorist we know of,” write Wright and Schaffner (2002, 367), “has. . . explicitly challengedSchattschneider’s . . . proposition.”
2. Other arguments focus on the danger that office holders, once elected, will attempt to stayin office unconstitutionally or otherwise tyrannize their subjects. In this view, parties’ ambitionfor office serves as an essential check on incumbents’ authoritarian tendencies, especially in weaklyconsolidated democracies (see, e.g., Levitsky and Cameron 2003, on Peru under Fujimori). Withoutparties, the logic goes, opposition to power-hungry incumbents will be fragmented and susceptible tocooptation. See also Ferejohn (1986) on collusion between incumbents and challenges in one-partystates.
3. Snyder and Ting (2003) provide the most compelling formalization of the argument that lack ofinformation inhibits representation. They show that increases in the probability that voters observerepresentatives’ behavior in office leads to greater median convergence and a stronger correlationwith constituency preferences on the part of representatives.
4. I use representation in the sense described by Pitkin (1967, 209): “acting in the interest ofthe represented, in a manner responsive to them.” This is roughly equivalent to what Dahl andothers mean by the term responsiveness. As I discuss below, however, I reserve the latter term fora specific empirical measure of representation: the expected difference in representatives’ behaviorassociated with a one-unit difference in constituency preferences (Achen 1978).
3
representation as partisan elections. To support this argument, I re-examine thecanonical example of the pathologies of one-party politics: the American South inthe early to mid-20th century (Key 1949).5
Focusing on congressional politics between the 1930s and early 1960s, I arguethat electoral competition in Democratic primaries created a “selectoral connection”for Southern members of Congress (MCs), inducing them to represent the (white)eligible electorate. I show empirically that Southern senators and representatives,typically considered to have represented only the economic elite, were in fact broadlyin step with the evolving policy preferences of Southern white public as a whole. Inaddition, Southern senators were no less responsive to their white constituents thanwere Northern senators to their constituents.
My explanation for these surprising results invokes several features of the largerinstitutional context. First, although Southern MCs did not face partisan competi-tion, they still operated in a partisan environment once in Congress. Southern MCswere forced to take positions in the low-dimensional space of national politics, givingvoters better information about their ideological positions than would have been thecase had Congress not been organized along partisan lines (cf. Wright and Schaffner2002). National partisan divisions also trickled down into Southern primary cam-paigns, leading candidates to mimic the ideological positions taken by Democratsand Republicans elsewhere. Finally, representation was enhanced by the relativelyhigh salience and competitiveness of Southern primaries. Consistent with this claim,Southern members of the House, whose primaries were somewhat less competitivethan senators’, appear to have been less responsive than both Southern senators andNorthern House members.
These findings contribute to two major literatures. The first concerns the place ofthe South in American political development. In line with the view of the one-partySouth as a set of “authoritarian enclaves” (Gibson 2012; Mickey 2014), scholars havetypically portrayed Southern MCs as agents of the Southern plantation elite and aslargely immune to popular pressures. Consequently, the large body of scholarshipdocumenting Southern MCs’ crucial role in structuring the American welfare stateand political economy attributes their pivotal actions almost exclusively to the pref-erences of the Southern elite.6 Although evidence of representation does not mean
5. As evidence for the one-party South’s canonical status, consider the fact that one recent reviewof parties’ role in democratic politics, Aldrich and Griffin (2010), devotes three of its thirteen pagesto the one-party South and discusses no other empirical examples (see also Aldrich 2011, 25–6,310–12).
6. See, for example, Katznelson, Geiger, and Kryder (1993), Quadagno (1994), Lieberman (1998),Alston and Ferrie (1999), Brown (1999), Manza (2000), Farhang and Katznelson (2005), andDomhoff and Webber (2011).
4
that the one-party South was a democracy, it does mean that our accounts of South-ern MCs actions must be rooted at least partly in the preferences of their whiteconstituents. Indeed, the possibilities for representation within the one-party systemhelp explain whites’ loyalty to—and thus the durability of—the South’s authoritarianregime.
My findings also shed light on parties’ place in democratic theory. First, theycast doubt the argument that because democracy requires representation and rep-resentation requires partisan competition, democracy requires partisan competition.There may be other reasons that democracy requires parties, but arguments thatrun through the mechanism of representation appear to be unsound. More generally,the evidence I present undermines the claim that partisan competition necessarilyenhances representation. When elections are regularly contested and voters haveaccess to low-cost information about candidates’ positions, as was especially true ofSenate primaries in the South, representation appears to work about as well as inpartsan elections. However, when these conditions do not pertain, as was more fullythe case in state and local politics in the one-party South, the conventional critiqueof one-party politics is much more likely to hold.
The remainder of this paper is organized as follows. I begin with an overview ofSouthern politics between the 1930s and the early 1960s, documenting Southern MCs’evolution from New Dealers to centrists as well as their persistent ideological diversityon questions of economics. I then describe the operation of Southern primaries, withan emphasis on how they induced Southern MCs to respond to the policy preferencesof the eligible electorate. Following that, I present empirical evidence for congres-sional representation in the one-party South. I first examine Southern Democrats’collective congruence with the economic preferences of the Southern white public,and then compare the dyadic responsiveness of state congressional delegations inthe North and South. The final empirical section analyzes responsiveness to theeconomic interests of the electorate, examining differences in Southern MCs’ respon-siveness between the House and Senate. The paper concludes with a discussion ofmy findings’ implications for democratic theory and American political development.
2 The One-Party South, 1930s–60s
For the first half of the 20th century, until the Second Reconstruction of the 1960s,the eleven states of the former Confederacy were ruled by a set of exclusionary one-
5
party enclaves (Gibson 2012; Mickey 2014).7 These regimes were established byDemocratic party leaders after several decades of often-violent struggle following theend of Reconstruction in 1876. The Democrats consolidated their hegemonic positionthrough the imposition of suffrage restrictions that disenfranchised blacks and manypoor whites, the core supporters of the opposing Republican and Populist parties(Woodward 1951; Kousser 1974).
The longer-term viability of these one-party Democratic regimes depended on an-other institutional innovation, the so-called “white primary,” which subjected Demo-cratic nominations to popular election. Until 1944, blacks could be legally barredfrom Democratic primaries, enabling whites to safely divide in the primary, minimiz-ing the risk of partisan defection in the (uncompetitive) general election (Kousser1984; Mickey 2008). The white primary was thus both a linchpin of one-party ruleand a forum for popular participation “through which the interests of low-incomewhites could be pressed” (Sanders 1999, 153). Democratic primaries remained themain site of electoral competition in the South until the early 1960s, when Republi-cans finally began to regularly contest and win general elections below the presiden-tial level (Black and Black 2002; Ansolabehere et al. 2010).
Notwithstanding the potential for popular participation offered by Democraticprimaries, most observers have instead emphasized the ways in which the one-partysystem inhibited representation—even of the eligible electorate. The classic state-ment of this view is offered by V. O. Key’s Southern Politics (1949). While acknowl-edging the diversity of and contestation within the one-party system, Key concludesthat intra-party factionalism was too localistic, personalistic, and transient to sub-stitute for partisan competition. Focusing on state government, Key (1949) makesseveral claims that have become enduring themes in the brief against one-party poli-tics (see especially chapter 14). The first is that without parties to structure politicalconflict and simplify political alternatives, voters lacked the information required forprospective policy-based choice between competing candidates. Further, the tran-sience and fragmentation of factional conflict, which made it difficult to distinguish“ins” from “outs,” inhibited even basic retrospective accountability. According toKey, the disorganized nature of “no-party” politics both inhibited responsiveness tovoter preferences and, by favoring the status quo, biased representation towards theinterests of the “haves.”
7. Like most works on Southern politics (e.g., Key 1949; Bullock and Rozell 2012), I define theSouth as the eleven states of the former Confederacy, though I sometimes discuss examples fromOklahoma and other Border states with substantial intra-party competition. Throughout the paperI exclude from my analysis the few Republican congressional districts in the South. I use the termNorthern as a synonym for non-Southern.
6
Scholars following in Key’s wake have tended to adopt a more extreme and lessnuanced version of his argument, which might be called an “elite dominance” per-spective. This perspective is exemplified by the following quotation from Domhoffand Webber (2011, 59):
Complete planter dominance through the Democratic Party was solidifiedby the “white primary” system introduced early in the twentieth century.This system maintained one-party white dominance while at the sametime providing a semblance of political choice and electoral competition,thereby allowing the dominant planter class to continue to profess itsallegiance to democratic principles. . . . With all due allowance for thepopulist-oriented stem-winders within the Democratic Party, this wasthe party of the Southern white rich, at whose core stood the plantationcapitalists.
2.1 Southern Democrats in Congress
Most studies of congressional politics in the one-party South adopt an elite dominanceperspective, treating Southern MCs as agents of planters and other low-wage employ-ers.8 In this view, lack of electoral competition gave Southern MCs “few incentivesto respond to whatever popular pressures did emerge” from the region’s “shriveled,conservative electorate” (Manza 2000, 309). As a consequence, Southern MCs, “freefrom the constraints of a conventional reelection imperative,” instead represented the“the interests of its economic and political elites” (Farhang and Katznelson 2005, 6;Katznelson, Geiger, and Kryder 1993, 284).9
While some analyses reduce Southern MCs to a reactionary bloc (e.g., Hofstadter1949), most careful scholars acknowledge the liberal and reformist impulses of the
8. Key (1949) devotes two chapters to a discussion of the South in Congress, in which he em-phasizes their ideological diversity on questions other than race. He notes that urban congressionaldistricts tended to send more liberal representatives to Congress, and that ideological differencesbetween senators tended to reflect their factional affiliations. But otherwise he has little to sayabout the sources of ideological variation among Southern MCs or their representational linkageswith voters.
9. Other works that more-or-less adopt an elite-dominance perspective on the one-party Southinclude Bensel (1984), Black and Black (1987, 2002), Quadagno (1994), Lieberman (1998), Alstonand Ferrie (1999), and Brown (1999), and Domhoff and Webber (2011). A particularly strong anddirect statement of this view is given by Domhoff (1996, 3–4), who stresses “the immediate, direct,and continuous role of large plantation owners in effecting federal policy through the Democraticparty and Congress. . . . [T]he planters were often in Congress, representing themselves.”
7
South as well as the ideological diversity within the South caucus.10 Ira Katznel-son, in various collaborations, has painted the fullest and most nuanced portrait ofSouthern Democrats’ role in national politics in the wake of the New Deal (Katznel-son, Geiger, and Kryder 1993; Farhang and Katznelson 2005; Katznelson and Mul-roy 2012; Katznelson 2013). Katznelson’s work emphasizes the trade-off SouthernDemocrats faced between securing federal aid for the underdeveloped South andinsulating the region’s racial political economy from external intervention.
The relative weight of these competing considerations changed over time. In theearly to mid-1930s, with their region still reeling from the economic devastation ofthe Depression, Southern MCs offered overwhelming support for Franklin Roosevelt’sNew Deal. In the 75th Congress (1937–38), however, the dangers posed by theNew Deal began to loom larger to Southern Democrats, who increasingly alliedwith Republicans in a “conservative coalition” to block or roll back liberal advances(Patterson 1967; Brady and Bullock 1980; Schickler and Pearson 2009). By themid-1940s, Southerners had reached a pivotal middle position in Congress, givingthem veto power over liberal and conservative initiatives and thus disproportionateinfluence over the American welfare state and political economy (Katznelson, Geiger,and Kryder 1993; Quadagno 1994; Lieberman 1998; Alston and Ferrie 1999; Farhangand Katznelson 2005). Southern Democrats’ moderate position on economics wasalso the primary cause of the anomalously low levels of partisan polarization in mid-century congresses.
Even many works that acknowledge Southern Democrats’ complex and conflictingmotivations typically treat them as a single “player” in congressional politics (e.g.,Katznelson, Geiger, and Kryder 1993). This has the unfortunate effect of downplay-ing the ideological diversity of the Southern Democratic Party, which, Key notes,“encompasses all shades of political attitude” on issues other than civil rights (1949,360). While nearly all Southern politicians supported Jim Crow, they varied in theweight they placed on regional racial autonomy relative to liberal economic policiesthat benefited many of their constituents.
Southern Democrats’ ideological diversity, as well as their evolution over time,can be visualized in Figure 1. For each congress between 1931 and 1962, this figureplots the ideological distributions within three groups of U.S. senators: NorthernDemocrats (blue), Southern Democrats (green), and Republicans (red), all of whomwere Northern. The ideological estimates are derived from a dynamic ideal-pointmodel that allows legislators to evolve flexibly between congresses. Only roll-callvotes related to social welfare and economic regulation, the core issues of the New
10. Only on civil rights issues did Southern MCs engage in the organization and coordinationnecessary for collective action (Finley 2008; Bloch Rubin 2014).
8
1931−1932 1933−1934 1935−1936 1937−1938
1939−1940 1941−1942 1943−1944 1945−1946
1947−1948 1949−1950 1951−1952 1953−1954
1955−1956 1957−1958 1959−1960 1961−1962
−2 −1 0 1 2 −2 −1 0 1 2 −2 −1 0 1 2 −2 −1 0 1 2Economic Conservatism
N. Dem. S. Dem. Repub.
Ideological Evolution in the U.S. Senate
Figure 1: Distribution of economic ideal points among Northern Democrats, South-ern Democrats, and Republicans in the U.S. Senate between the 72nd (1931–32) and87th (1961–62) congresses. Ideal points have been coded so that conservatives areon the right. Vertical lines indicate the mean ideal point in each group.
9
Deal, are used to estimate the model, so the ideal points can be interpreted ascongress-specific measures of senators’ opposition to New Deal liberalism.11
Figure 1 highlights two important patterns in Congress (analogous plots for theU.S. House tell a similar story). The first is Southern Democrats’ transformationfrom strong liberals in the 1930s to moderates by 1945–46 (79th Congress). Thismade Southern MCs collectively the pivotal player in congressional politics, a positionthey maintained into the 1960s (Katznelson and Mulroy 2012). But Figure 1 alsoillustrates the limitations of treating Southern Democrats as a collective actor, forthey remained ideologically diverse even as they became more conservative. Preciselybecause the pivotal voter was so often a Southern Democrat, the exact ideologicaldistribution within the Southern caucus usually determined the set of feasible policyoutcomes.
The ideological diversity within the Southern caucus is sometimes acknowledged,but it is almost never explained. This is partly because most accounts either treatSouthern MCs as ciphers for an undifferentiated elite, or else explain their diversity interms of idiosyncratic variation in Southern MCs’ personal policy preferences. Onlya few works have rooted their explanations in the “electoral connection” so centralto analyses of congressional behavior outside the one-party South (Mayhew 1974a).Fewer still have attempted to measure the preferences of Southern electorates andexamine their relationship with the behavior of Southern MCs.12
3 The Selectoral Connection in the One-Party South
In order for Southern MCs to have responded to popular preferences, there hadto have been a mechanism to induce them to do so. In democratic regimes, this
11. The ideal points were estimated using the dynamic Bayesian one-dimensional IRT modeldeveloped by Martin and Quinn (2002) and implemented in the R package MCMCpack (Martin,Quinn, and Park 2011). There are several reasons that I use these estimates rather than off-the-shelf estimates, such as first-dimension DW-NOMINATE scores (Poole and Rosenthal 2007).First, I wish to restrict the analysis to bills related to New Deal liberalism, which I do by usingonly roll calls in the “Government Management” and “Social Welfare” issue categories developedby Clausen (1973) and included in data provided by Poole and Rosenthal (1998, 2001). Second,publicly available DW-NOMINATE scores constrain ideal points to move linearly across congresses,which makes it difficult to detect rapid or non-monotonic ideological change. Third, estimating themodel with MCMC simulation facilitates the calculation of uncertainty estimates for auxiliaryquantities, such as the median in each congress (Clinton and Jackman 2009). See Caughey andSchickler (2014) for further discussion of these issues.
12. Three partial exceptions are Flinn and Wolman (1966), Mayhew (1966), and Fleck (2002),who proxy for the electorate’s in various ways.
10
mechanism is elections, which (potentially) allow voters to select representative can-didates and sanction unrepresentative incumbents (Fearon 1999). The combinationof prospective selection and retrospective accountability tends to align office-holdersincentives with the preferences of their electorates. Given the lack of competitionin general elections, Democratic primaries are the only plausible candidate for sucha mechanism. Were they up to the task? That is, did they provide primary voterswith meaningful and identifiable alternatives, and were incumbents held electorallyaccountable once in office?
In this section, I present evidence that Democratic primaries did indeed create a“selectoral connection” between MCs and voters, providing incentives for the formerto be responsive to the latter.13 After first showing that the electorate extended wellbeyond the economic elite, I demonstrate that competition in Democratic primarieswas frequent and often presented voters with clear ideological choices. Finally, Idiscuss the evidence for responsiveness in Southern primaries, showing that ideolog-ical turnover among Southern MCs often corresponded to changes in voters’ policypreferences.
3.1 Participation
Following Dahl (1971), we can distinguish between two dimensions of democracy:inclusiveness (the scope of political participation) and contestation (the extent ofpermissible political competition). The one-party South exhibited limitations onboth of these dimensions. Most blatantly, suffrage restrictions excluded many whitesand nearly all blacks from political participation (Kousser 1974; Alt 1994). For itspart, lack of partisan competition was maintained not only by the Democratic loyal-ties of enfranchised whites, but also by a variety of legal and extralegal mechanismsthat raised the costs of partisan opposition (Mickey 2008, 149; see also Valelly 2004on the barriers to party-building in South).
Distinguishing inclusiveness from contestation is essential to identifying the con-sequences of lack of party competition per se. All works that have used the caseof the one-party South to argue that party competition is necessary in democraticregimes have relied on the assumption that it is possible to separate the effects ofone-partyism from those of disenfranchisement. The central link in this line of rea-soning is the claim that lack of party competition denied representation even to theenfranchised population.
One way the elite-dominance perspective could be reconciled with representation
13. I borrow the selectoral connection from Manion (2014), who combines Mayhew’s (1974) elec-toral connection with the term selectorate from Bueno de Mesquita et al. (2003).
11
of the eligible electorate is if political participation was so restricted that only theeconomic elite were included. If this was so, the selectorate of the one-party Southwas so small as to be oligarchical. Moreover, such a small set of principals must havebe able to monitor representatives very closely, inducing a close agency relationshipand a high degree of representation elite preferences (cf. Gailmard and Jenkins 2009).
Data on voter turnout, however, suggest that the electorate extended well beyondthe economic elite. In presidential elections between 1932 and 1948, turnout in theSouth averaged about 25% of the voting-age population (VAP); in 1952–60, the figurewas almost 40%.14 As a point of comparison, presidential turnout in the South roseto a peak of 50% immediately after the 1965 Voting Rights Act and remained roughlyat that level though the end of the century (McDonald and Popkin 2001, 969, Table3). Over the 1932–60 period, the number of presidential voters in Southern generalelections ranged between one-third and one-half of the white population.15
The congressional primary electorate was usually but not always a bit smallerthan the presidential electorate. Turnout in contested House and Senate primaries av-eraged 17–23% of VAP in years before 1945 and 24–27% afterwards.16 Since primaryturnout depended on election-specific conditions, presidential turnout is probably amore accurate (though still conservative) measure of the eligible primary electorate.17
White turnout, which was depressed by apathy (Ewing 1953, 3–4) and free-riding(Alt 1994, 359) as well as by suffrage restrictions, was inherently more elastic thanturnout among blacks, for whom suffrage barriers were almost insuperable.
3.2 Competition
A core attribute of political parties is that they contest elections in an effort to gaincontrol of government. Effective representation may not require that elections becompetitive in equilibrium,18 but without the potential for competition, “the right
14. Calculated from county-level demographic data from the U.S. Census and electoral data fromLeip (2013).
15. African Americans, who constituted about a quarter of the South’s population, were almostcompletely disenfranchised until 1944, and they remained a small presence in Southern electionsuntil after 1965 (Bartley 1995, 172).
16. Calculated from data provided by James Snyder on behalf of his collaborators (Ansolabehereet al. 2010).
17. The eligible electorate is a more relevant quantity than the actual number of voters becausepoliticians tend to act in light of the anticipated response of potentially mobilizable constituencies(Arnold 1990).
18. For example, in Besley and Coate’s (1997) “citizen-candidate” model of representative democ-racy, the policy outcome in an equilibrium in which only a single candidate runs for office is alwaysthe ideal point of the median voter (given a one-dimensional policy space).
12
to ‘participate’ is stripped of a very large part of [its] significance” (Dahl 1971,5). Did Democratic congressional primaries compensate for the almost total lack ofcompetition in Southern general elections?19
To a substantial degree they did (Ansolabehere et al. 2010). Over the 1932–60period, 45% of Democratic House nominations and 56% of Senate nominations werecontested. Primary competition declined a bit in the 1950s, but primaries remainedat least as important as general elections as sites of political competition. Over30% of Southern members of Congress experienced at least two contested primaryvictories over their career. Even Sam Rayburn of Texas, the powerful and long-serving Speaker of the House, rarely saw an election year pass without a primarychallenge (Hardeman and Bacon 1987).
Primary opposition was not just perfunctory. About 19% of contested congres-sional primaries were decided by a margin less than 10 percentage points, and 36%were decided by less than 20 points.20 Incumbents running for reelection were de-feated in 7% of all Southern primaries in the 1930s, 3.5% in the 1940s, and 2% in the1950s—incumbent defeat rates comparable to many contemporary House elections(Ansolabehere et al. 2010, 197). In the Senate, the incumbent defeat rate was muchhigher, closer to 30%: between 1920 and 1954, 18 Southern senators were defeatedfor renomination.21
Was this competition meaningful? That is, was is structured along ideologicallines that were inelligible to most voters? Many scholars have emphasized that inthe absence of party labels, Southern primary campaigns often revolved around othersalient characteristics of the candidates, such as a flamboyant personality. But pri-mary candidates also frequently took ideologically opposing positions. As Key (1964,441) observes, Southern congressional primaries in this period “tend[ed] to take onthe tone of contests between Democrats and Republicans elsewhere.” As it did in
19. General elections in the one-party South tended to be sleepy affairs in which the Democraticnominee cruised to victory. Even as Republicans made inroads at the presidential level in the 1950s,competing for lower-level offices in the South far more daunting, given that the party totally lackedexperienced candidates or an institutional base in the region (Black and Black 2002). Notwithstand-ing Republican takeovers of a few isolated congressional districts, over three-quarters of DemocraticHouse candidates between 1950 and 1960 won office with more than 80% of the vote. Through the1960 election, Democrats continued to compose well over 90% of Southern House members and100% of Southern senators.
20. When elections that were uncontested or for which data are missing are included in the de-nominator, the percentages fall to 8% and 17%, respectively. As a point of comparison, in the 35years after the 1974 publication of David Mayhew’s famous “vanishing marginals” article, about15% of non-Southern House elections were decided by less than 10 points.
21. The figures for the Senate, taken from Key (1964, 441), include Oklahoma and exclude senatorsappointed to office.
13
national politics, the New Deal provided a focal point for ideological conflict andsharpened policy-based divisions within the one-party South. Examples of closelycontested and ideological Senate primaries include Bilbo vs. Stephens (MS, 1934),Fountain vs. Bailey (NC, 1936), Hill vs. Heflin (AL, 1938), Johnson vs. O’Daniel(TX, 1941), Johnston vs. Thurmond (SC, 1944), Sparkman vs. Boykin (AL, 1946),Long vs. Kennon (LA, 1948), Pepper vs. Smathers (FL, 1950), Gore vs. McKellar(TN, 1952), McMath vs. McClellan (AR, 1954), and Yarborough vs. Blakley (TX,1958), all of which pitted a New Deal liberal against a substantially more conserva-tive opponent. Campaign appeals, media coverage, and interest-group endorsementshelped convey this information to voters, who also could also rely on heuristics suchas a candidate’s hometown.22
3.3 Responsiveness
Congressional candidates may have taken ideologically opposed positions, but didvoters have enough information to perceive differences between candidates and makeideological choices? Grynaviski (2004) provides evidence that Southern primary vot-ers were able to coordinate on the optimal number of viable candidates, especially inhigher-salience contexts, such when an incumbent was running for reelection. Moredirect evidence for ideological accountability in Southern primaries is provided bythe numerous examples of Southern MCs denied renomination when the preferencesof the electorate changed.
Such preference shifts occurred for two reasons. One was the mobilization ofnew constituencies into the electorate. In 1948, for example, CIO-backed HendersonLanham mobilized newly unionized textile workers and unseated anti-labor Georgiacongressman Malcom Tarver (Brattain 1997).23 Returning WWII veterans played asimilar role in the 1948 Tennessee Senate primary, helping the liberal Estes Kefauverdefeat Memphis machine boss E. H. Crump (Cobb 2010, 67).24
More commonly, shifts in the ideological mood of the Southern public led toturnover in the region’s congressional delegations. “In a period of conservatism,”
22. Key (1949) notes the prevalence of campaign-specific “friends and neighbors” alignments inprimary elections. But he also discusses more durable and substantive geographic cleavages, mostcommonly between the converative black belt and the populist upcountry (see also Nixon 1946).
23. Brattain (1997, 96) writes, “For all its limits, the South’s one-party system proved flexibleenough to allow [textile union] members to use it as a vehicle to express their class as well astheir white racial interests.” See also Albert Rains’s 1944 defeat of Alabama Representative JosephStarnes with the aid of union and CIO-PAC support (Rosenfarb 1944; Norrell 1991, 228).
24. In general, WWII veterans proved to be a destabilizing force in Southern politcs, but as oftenin a conservative direction as in a liberal one (Brooks 2000).
14
noted one contemporary observer, “the one-party South responds by electing conser-vative Democrats while the farm belt responds by electing reactionary Republicans”(Carleton 1951, 226). Years with a national Republican swing, such as 1938, sawthe primary defeats of Southern liberals like Rep. Maury Maverick of Texas andRep. Norman Hamilton of Virginia (Weiss 1971; Sweeney 1999).25 Many of theseliberals had been swept into office in the early to mid-1930s, a period of liberal as-cendency when a number of anti–New Deal incumbents were defeated by more liberalchallengers.26
The threat of such challenges prompted an ancipatory response from incumbents.Some responded by retiring pre-emptively from office. Coode (1971, 19) describesthe 1936 retirement decision of Bryant Castellow, a conservative House member fromGeorgia, as follows:
[T]ired of public office, weary of facing a campaign against a serious con-tender he had defeated by only 2,000 votes two years before, apparentlyhaving lost whatever earlier enthusiasm his voting record suggest that hehad for the New Deal, and unable to stop the march of the new order,Castellow declined to run for re-election.
Pro-labor North Carolina Representative John Folger made the same decision aftersqueaking out renomination in 1946.27 Many other incumbents responded to shiftingideological winds by adapting their roll-call voting. Indeed, contrary to the conven-tional finding that ideological change in Congress occurs overwhelmingly throughreplacement (Poole and Rosenthal 2007; but see Stimson, MacKuen, and Erikson1995), adaptation was at least as important as replacement as a mechanism forSouthern MCs ideological migration in this period.28
25. 1938 was also the year of FDR’s failed “purge” of conservative Democrats, most of themSouthern (Dunn 2010).
26. In 1934, for example, Senator Hubert Stephens of Mississippi, a lukewarm supporter of theearly New Deal, tried in vain to tie himself to Roosevelt but was defeated for renomination by thepopulist Theodore Bilbo—the sort of challenge on the left that kept otherwise wary Southernersloyal to FDR (Morgan 1985, 64; Swain 2001). One senator who did not heed the signs of this defeatwas Oklahoma’s Senator Thomas Gore, a long-time incumbent and one of the most outspokenopponents of the New Deal, who was overwhelmingly defeated in the 1936 Democratic primary in(Patterson 1967, 22–4; Dangerfield and Flynn 1936).
27. Taking his near-defeat as a sign of increased conservatism in his district, Folger retired andwas succeeded by his more conservative challenger from two years previous (Christian 1976).
28. In most congresses, the average difference between exiting and entering members of the South-ern caucus was about as large as the the inter-congress differences among continuing members.Since there were many more of the latter, adaptation accounts for a much larger share of SouthernDemocrats’ ideological change between congresses.
15
In short, primary competition was far from universal, but it was common enoughto pose a genuine threat to Southern MCs, and competition did not have to beobserved to be effective. Further, while candidates typically (though not always) ranindividualized campaigns rather than affiliating with a coherent faction, the conductand outcomes of primary campaigns was not simply random. Rather, they oftenreflected shifts in public mood or the mobilization of new constituencies.
4 Statistical Evidence for Representation
Democratic primaries appear to have created a selectoral connection between South-ern MCs and the eligible electorate, but did this connection result in meaningfulrepresentation? That is, did MCs act in the interests of their voting constituents, ina manner responsive to their expressed preferences (Pitkin 1967, 209)? This sectionexplores this question from several angles, beginning with an examination of publicpreferences towards New Deal liberalism.
4.1 New Deal Liberalism in the Public
Evaluating the representativeness of Southern MCs first requires measures of thepreferences of their constituents. To construct such measures, we rely on commercialpublic-opinion polls fielded between 1936 and 1952.29 In these years, George Gallup,Elmo Roper, and other pollsters fielded hundreds of polls that collectively surveyedover one million Americans. These polls included hundreds of policy-related ques-tions, most of them on the issues of economic regulation and social welfare raisedby the New Deal and Fair Deal.30 These polls thus provide the raw material for
29. Only recently have these data become easily accessible to researchers, thanks to a team ledby Adam Berinsky, Eric Schickler, and Jasjeet Sekhon, who cleaned, standardized, and weightedthe poll samples. This work was funded by two grants from National Science Foundation, PoliticalScience Program: SES-0550431 (Berinsky and Schickler, “Collaborative Research: The Ameri-can Mass Public in the 1930s and 1940s,” 2006–2010) and SES-1155143 (Berinsky, Schickler, andSekhon, “Collaborative Research: The American Mass Public in the Early Cold War Years,” 2012–2014). For further details on the project and the data, see Berinsky (2006) and Berinsky et al.(2011).
30. The abundance of left–right domestic-policy questions from the 1936–52 period stands insharp contrast to the dearth of such questions included in surveys conducted between 1952 and theearly 1960s (Erskine 1964). It is difficult to say whether pollsters’ declining interest in questionsseparating liberals and conservatives is more a cause or a consequence of the relatively consensual,non-polarized politics of the 1950s.
16
a measure of public support for New Deal liberalism analogous to the ideal-pointestimates for MCs presented earlier.
Despite this wealth of information on public opinion, constructing a mass-levelmeasure of New Deal liberalism presents three challenges. First, because respondentswere selected using quota sampling rather than probability sampling, the poll samplesare not representative of the American public (Berinsky 2006). Second, though thepoll samples are comparatively large (2,000–3,000 respondents), they are still notlarge enough for reliable estimation of subnational opinion below the regional level.Third, because each poll rarely included more than a couple of policy questions, ideal-point techniques such as IRT models cannot be applied to individual respondents.
Previous works have dealt with these problems by weighting each poll to match re-gional population targets, which alleviates the sampling bias, and then analyzing theweighted poll-specific question marginals (e.g., Berinsky 2009; Schickler and Caughey2011). An alternative approach, better suited to our purposes, is the Bayesian group-level IRT model developed by Caughey and Warshaw (2013). In essence, their modelestimates the average (latent) liberalism in groups defined by geographic or demo-graphic characteristics, alleviating the problem of sparse subsamples by smoothingthe group estimates with a dynamic hierarchical model. The resulting group esti-mates can then be weighted in proportion to their composition of the population.31
This approach thus uses information from many question responses spread acrossmany polls to derive estimates of the aggregate distribution of liberalism.
We use the Caughey-Warshaw model to estimate mass-level liberalism in eachyear between 1936 and 1952. Like the congressional IRT estimates, the mass-levelestimates are based only on policy questions related to New Deal liberalism, whichstill yields over 450 distinct question series. Figure 2 plots the yearly hierarchicalcoefficient estimates from a model that defines groups by the variables Farm, PhoneOwner, Professional, Urban, and the interaction of Southern and Black.32 As thisfigure illustrates, New Deal liberalism is negatively associated with indicators ofsocioeconomic status, with residence in rural agricultural areas, and with being white.Regional patterns are not readily apparent in this graph, but they become clearerwhen we examine the weighted group estimates.
In our analysis of representation, we rely on estimates of liberalism in two sets of
31. In addition, the group-level question marginals used to estimate the model can also beweighted. This framework’s combination of hierarchical smoothing and post-estimation weight-ing is similar to the multilevel regression and poststratification method described by Gelman andLittle (1997) and Park, Gelman, and Bafumi (2004).
32. Note that these are the structural coefficients from a model predicting group liberalism, not theestimates of mean group liberalism themselves. Within each group, respondents have been weightedto be balanced between men and women, the former being overrepresented in the samples.
17
Non
−S
outh
ern
Whi
teS
outh
ern
Whi
teN
on−
Sou
ther
n B
lack
Sou
ther
n B
lack
Farm
Pho
neP
rofe
ssio
nal
Urb
an
−0.
4
0.0
0.4
0.8
−0.
4
0.0
0.4
0.8
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
Year
Coefficient Estimate (1 and 2 SE)
Fig
ure
2:H
iera
rchic
alpre
dic
tors
ofgr
oup
lib
eral
ism
.T
he
top
row
plo
tsth
ees
tim
ated
inte
rcep
tsof
the
four
mutu
ally
excl
usi
veSouth×Race
grou
ps.
The
bot
tom
row
plo
tsth
eco
effici
ents
for
four
dem
ogra
phic
pre
dic
tors
.A
sa
poi
nt
ofco
mpar
ison
,th
est
andar
ddev
iati
onof
lib
eral
ism
acro
ssal
lre
spon
den
tsis
esti
mat
edto
be
0.78
.
18
population groups.33 The first set of groups is defined by the interaction of the vari-ables South, Black, and Presidential Vote (Democratic, Republican, and Other/DidNot Vote). These estimates are presented in Figure 7 in the Supplementary Ap-pendix. The second set of groups is defined by the interaction of State and Black.We use the State×Black estimates to derive the average liberalism in the electorateof each state. In the non-South, this entails weighting race-specific group estimatesaccording to the racial composition of the states’ citizens. In the South, we definethe electorate as the white public, dropping Southern blacks from the analysis.34 To-gether with the ideal-point estimates discussed earlier, these estimates of mass-levelliberalism form the empirical basis for the analysis of representation that follows.
4.2 Collective Representation
We begin with an examination of Southern senators’ and representatives’ collectiverepresentation of the Southern white public.35 In the aggregate, were SouthernMCs’ roll-call records on economic issues congruent with the expressed preferencesof Southern whites? Since the congressional and mass-level estimates of liberalismare not on the same scale, we cannot directly measure their spatial proximity orcongruence (Achen 1978; Matsusaka 2001).36 We can still learn a great deal, however,
33. For both sets of estimates, we weight the raw survey responses to adjust for class and genderbiases in the samples. The population targets for these variables were derived from census andother data using the interpolation method described in Caughey and Wang (2014).
34. Southern blacks were severely undersampled—or not sampled at all—in the early polls, mostof which sought to make their samples representative of voters, not of the American public at large.Although many polls include a question on retrospective presidential vote, we do not use this variableto define the voting population because (1) voting is highly overreported and (2) this definitionwould preclude us from weighting the model estimates since the demographic characteristics ofvoters is not known. That said, there appears to have been very little difference in liberalsim betweenSouthern whites who reported having voted for the Democratic presidential candidate and non-voting Southern whites. However, presidential Republicans, who made up an increasing proportionof the Southern whites after 1948, were substantially more conservative (this contrasts with Shaferand Johnston’s claim that Republican vote choice was unrelated to social welfare attitudes in the1950s; Shafer and Johnston 2006, 28–9). The conservative impact of rising Republican shares wascounterbalanced by the fact that Republicans were less likely to participate in Democratic primariesand by the contemporaneous enfranchisement of a modest number of (more liberal) Southern blacks.In short, the ideological difference between the white public and the Democratic primary electoratewere probably minimal, especially in the years 1937–52.
35. The term collective representation is taken from Weissberg (1978). Our usage differs slightlyfrom his in that he applies it to Congress as an institution, whereas we apply it to a collectivity(Southern Democrats) within that institution.
36. For example, Achen’s analysis of House members’ proximity to their constituents relies on theassumption that the Guttman issue scales for representatives and citizens have the same scale once
19
by comparing relative positions of different groups of citizens and legislators.Figure 3 compares trends in liberalism among Northern Democrats, Southern
Democrats, and Republicans in three venues: the white public, the U.S. House,and the U.S. Senate.37 Both at the mass level and in Congress, Southern Democratsbegan the 1937–52 period as enthusiastic New Dealers, almost as much so as NorthernDemocrats. By the mid-1940s, Southern Democrats in all three venues had movedto a moderate position about halfway between Northern Democrats and NorthernRepublicans.38
The similarities across the three venues suggest that Southern MCs were broadlyin step with the Southern white public. Moreover, there is little evidence that theSouthern whites simply followed the changing views of their congressional represen-tatives (cf. Lenz 2012). Indeed, the Southern white public’s shift to the ideologicalcenter in 1937–42 clearly preceded the corresponding shift among Southern sena-tors. Thus, while we cannot rule out the possibility that both shifts were driven bysome third factor, such as the changing views of the Southern elite, these figuresprovide suggestive evidence of dynamic representation on the part of Southern MCs(Stimson, MacKuen, and Erikson 1995).
4.3 Responsiveness to Economic Policy Preferences
Southern MCs appear to have been broadly in step with the Southern white public,but were they responsive to ideological differences between their electorates? Wasresponsiveness greater in the non-South, and did it differ between the House andSenate? These questions present something of a hard test for my argument becauseresponsiveness is a hallmark of competition between polarized parties—even to thepoint of “leapfrog representation” (Bafumi and Herron 2010). In other words, it ispossible that Southern MCs were collectively congruent with their electorates withoutbeing individually responsive to them.
they have been standardized (see Achen 1978, 485).37. We define partisanship in terms of retrospective presidential vote since this is the most fre-
quently available indicator. Although blacks could vote in the North, they are excluded for consis-tency’s sake from the mass-level estimates in the North as well as the South. Including Northernblacks in the analysis slightly increases the relative conservatism of Southern white Democrats. Thesame is true if Southern white Republicans are included in the analysis.
38. Another notable feature of this graph is the partisan convergence in the early war years andrenewed polarization after the war. This pattern may be the result of wartime bipartisanship,which temporarily muted partisan divisions over the New Deal. It should also be noted that sincepartisanship is defined in terms of retrospective presidential vote, the composition of partisan groupschanges after each election.
20
●
●● ●
● ● ●●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
−0.6
−0.3
0.0
0.3
1937−38 1939−40 1941−42 1943−44 1945−46 1947−48 1949−50 1951−52
Rel
ativ
e Li
bera
lism
(95
% C
I)
●
●
●
N. Dem.
S. Dem.
N. Rep.
White Public
● ●
●●
●
●● ●
● ●
●
●
●
●
●
●
●●
●●
●
●
●●
−0.5
0.0
0.5
1937−38 1939−40 1941−42 1943−44 1945−46 1947−48 1949−50 1951−52
Libe
ralis
m (
95%
CI)
●
●
●
N. Dem.
S. Dem.
N. Rep.
U.S. Senate
●
● ●
●
●●
● ●
●
●●
●
●
●
● ●
● ● ●●
● ●● ●
−1.0
−0.5
0.0
0.5
1.0
1937−38 1939−40 1941−42 1943−44 1945−46 1947−48 1949−50 1951−52
Libe
ralis
m (
95%
CI)
●
●
●
N. Dem.
S. Dem.
N. Rep.
U.S. House
Figure 3: Trends in the economic liberalism of the white public (top), U.S. Senate(middle), and U.S. House (bottom), by region and party, 1937–52.
21
AL
AR
AZ
CA
CO
CT
DE
FL
GA
IAID
ILIN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
NDN
E
NHN
J
NM
NV
NY
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL A
RAZ CA
CO C
T DE
FL
GAIA ID
IL IN
KSK
YLA M
A
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM
NV
NY
OH
OK
ORPA
RI
SC S
D
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
AR
AZ
CA
CO
CT
DE
FL
GA
IA IDIL
INK
S
KY LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA RI
SC
SDTN
TX
UT
VA
VT
WA
WI
WV
WY
AL AR
AZ
CA
CO
CT
DE
FL
GA
IAID
IL
IN
KS
KY LA
MA
MD
ME M
IMN
MO
MS
MT
NCND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL AR
AZ
CA
CO
CT
DE
FL
GA
IA
ID
IL
INK
S
KY
LAM
AM
D
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM NV
NY
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
AR
AZ
CA
CO
CT
DE
FL
GA
IAID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NEN
H
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
AR
AZ
CA
CO
CT
DE
FL
GA IAID
IL
IN
KS
KY
LA
MAM
D
ME M
I
MN
MO
MS
MTN
C
ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC S
D
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
ARAZ
CA
CO
CT
DE
FL
GA
IA
ID
IL
IN
KS
KY
LA
MA
MD
MEMI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM N
V
NY
OH
OK
OR PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
1937
−19
3819
39−
1940
1941
−19
4219
43−
1944
1945
−19
4619
47−
1948
1949
−19
5019
51−
1952
−1.
0
−0.
5
0.0
0.5
1.0
1.5
−1.
0
−0.
5
0.0
0.5
1.0
1.5 −
0.8
−0.
40.
00.
4−
0.8
−0.
40.
00.
4−
0.8
−0.
40.
00.
4−
0.8
−0.
40.
00.
4Li
bera
lism
of S
tate
Pub
lic (
Sou
ther
n B
lack
s E
xclu
ded)
Liberalism of Senate Delegation
aa
Non
−S
outh
Sou
th
Fig
ure
4:Sen
ator
ial
resp
onsi
venes
sto
mas
spre
fere
nce
s,by
regi
onan
dco
ngr
essi
onal
term
.T
hex
-axis
plo
tsth
ees
tim
ated
mea
nec
onom
iclib
eral
ism
inth
est
ate
public,
aver
aged
over
the
two
year
sin
each
congr
essi
onal
term
.T
hey-a
xis
indic
ates
the
mea
nec
onom
iclib
eral
ism
ofth
est
ate
Sen
ate
del
egat
ion.
The
opin
ion
esti
mat
esfo
rSou
ther
nst
ates
do
not
incl
ude
bla
ckci
tize
ns.
Reg
ress
ion
lines
are
wei
ghte
dby
the
pre
cisi
onof
the
opin
ion
esti
mat
es.
22
The evidence, however, suggests otherwise. Consider Figure 4, which plots therelationship between the average liberalism of state publics and Senate delegations inthe 75th (1937–38) through 82nd (1951–52) congresses. The estimates for Southernpublics include only white citizens, which still represents an expansive definition ofthe Southern electorate.39 Nevertheless, the relationship between Southern publicsand their senators is clearly positive in every congress except the 77th (1941–42), amoment of ideological flux in Southern opinion.
Moreover, there are no obvious differences senatorial responsiveness between theSouth and non-South. In both Congress and the public, the ideological range acrossstates is smaller in the South than the non-South, but the regression slopes in the tworegions are approximately parallel. Indeed, the biggest regional difference is that by1939–40, Southern senators look anomalously liberal relative to their state publics,who began to cluster at the conservative end of states. The Senate delegation ofLouisiana, for example, is nearly a standard deviation more liberal than states witha similarly conservative mass public, such as Iowa or South Dakota.40
Figure 5 replicates the same analysis but for state delegations to the U.S. Houseof Representatives. Since public liberalism must be estimated at the state level, thematch between House members and their constituencies is not perfect, especially inthe South, where congressional districts were highly unequal in population. Still,in most years, Southern states with more liberal white publics tended to have moreliberal House delegations. The relationship between public and House liberalism,however, appears to have been stronger and more consistent in the non-South.
Formal statistical analysis confirms these visual patterns. Table 1 presents theresults of a model of responsiveness in congresses between 1936 and 1952.41 The
39. Including blacks in the Southern public barely alters the results, perhaps because there aretoo few black respondents in the data for precise estimation of black opinion. Including all whites inthe “public” is an expansive definition of the Southern potential electorate because most Southernwhites did not vote (to a lesser degree, this is also true of the non-South).
40. “The rest of the world,” observes Carleton (1951, 226),
insists on comparing the South with the records of Northern Democrats, largely fromthe industrial sections of the East and West coasts, and not with the farm belt. Peopleoutside the South forget that in a period of conservatism in the agricultural sectionsthe one-party South responds by electing conservative Democrats while the farm beltresponds by electing reactionary Republicans. But Republicans are expected to beconservative and no one is surprised at the farm belt, but Democrats are expected tobe liberal and when Southern Democrats are not liberal this fact is made cause forcomment. . . . And yet. . . the voting records of the South in Congress [are] more liberalthan those of the farm belt.
41. This analysis entails two kinds of averaging. First, within each congress, we average the es-
23
AL
AR
AZ
CA
CO
CT
DE
FL
GA
IA
ID
ILIN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND N
E
NHN
J
NM
NV
NY
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL A
RAZ CA
CO C
T DE
FL
GA IAID
IL IN
KSK
Y
LA MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR PA
RI
SC S
D
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
AR
AZ
CA
CO
CT
DE
FL
GA
IAID
IL
IN
KS
KYLA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PARI
SC
SDTN
TX
UT
VA
VT
WA
WI
WV
WY
AL AR
AZ
CA
CO
CT
DE
FL
GA
IA
ID
IL
IN
KS
KYL
A
MA
MD
MEM
I MN
MO
MS
MT
NC ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL AR
AZ
CA
CO
CT
DE
FL
GA
IA
ID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NMNV
NY
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
AR
AZ
CA
CO
CT
DE
FL
GA
IA ID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE N
H
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
AR
AZ
CA
CO
CT
DE
FL
GA IAID
IL
IN
KS
KY
LA
MAM
D
MEM
I
MN
MO
MS
MT N
C
ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC S
D
TN
TX
UT
VA
VT
WA
WI
WV
WY
AL
ARAZ
CA
CO
CT
DE
FL
GA
IA
ID
ILIN
KS
KY
LAM
AM
D
MEMI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NMN
VN
Y
OH
OK
ORPA
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
WV
WY
1937
−19
3819
39−
1940
1941
−19
4219
43−
1944
1945
−19
4619
47−
1948
1949
−19
5019
51−
1952
−2
−101
−2
−101 −
0.8
−0.
40.
00.
4−
0.8
−0.
40.
00.
4−
0.8
−0.
40.
00.
4−
0.8
−0.
40.
00.
4Li
bera
lism
of S
tate
Pub
lic (
Sou
ther
n B
lack
s E
xclu
ded)
Liberalism of State Delegation in U.S. House
aa
Non
−S
outh
Sou
th
Fig
ure
5:H
ouse
resp
onsi
venes
sto
mas
spre
fere
nce
s,by
regi
onan
dco
ngr
essi
onal
term
.T
hex
-axis
plo
tsth
ees
tim
ated
mea
nec
onom
iclib
eral
ism
inth
est
ate
public,
aver
aged
over
the
two
year
sin
each
congr
essi
onal
term
.T
hey-a
xis
indic
ates
the
mea
nec
onom
iclib
eral
ism
ofth
est
ate
Hou
sedel
egat
ion.
The
opin
ion
esti
mat
esfo
rSou
ther
nst
ates
do
not
incl
ude
bla
ckci
tize
ns.
Reg
ress
ion
lines
are
wei
ghte
dby
the
pre
cisi
onof
the
opin
ion
esti
mat
es.
24
model includes separate congress-chamber-region intercepts and a three-way interac-tion between Non-South, Senate, and Citizen Liberalism, clustering standard errorsby state. The analysis has low power because there are just eleven Southern states,but there is robust evidence of responsiveness in the one-party South in both theHouse (represented by the main effect of Citizen Liberalism) and the Senate (themain effect plus its interaction with Senate).42
There is also suggestive evidence of differences between the House and Senate.First, Southern senators are estimated to be more responsive than Southern represen-tatives, though the difference is not statistically significant.43 Second, as indicatedby the estimate for Non-South × Citizen Liberalism, responsiveness among Housemembers appears to have been stronger outside the South (p = 0.08). In the Senate,however, there is no evidence that responsiveness differed across regions—in fact, thepoint estimate for the slope difference between regions is almost exactly 0.
In short, Southern states where the white population was more supportive ofNew Deal liberalism tended to be represented by more liberal House and Senatedelegations. The evidence thus suggests that Southern MCs were not only collectivelyin line with the Southern white public, but also responsive to within-region differencesin white preferences. To the extent that Southern Democrats were out-of-step withtheir constituents, they appear to have been more liberal than the white public,at least relative to the non-South. We should not make too much of this “liberalbias,” which is almost certainly explained by the pull of party loyalty on the South’soverwhelmingly Democratic congressional delegation. But it is worth noting that thisbias is inconsistent with Key’s (1949) claim that the one-party system benefitted the
timated liberalism of senators in each state (and do the same for representatives). This yields thevariable MC Liberalism. Second, we estimate each state’s average liberalism over the course of eachcongress by averaging the state’s estimated mean liberalism across the two years in that congress.This yields the variable Citizen Liberalism. We calculate the standard error of the derived aver-ages under the assumption of independent measurement error across units (this will overestimatemeasurement error if the errors are positively correlated, as they probably are). The public opinionestimates for the 74th Congress are derived from a handful of polls conducted in the last monthsof 1936, and as a result the opinion estimates for this congress are fairly imprecise. To account formeasurement error in the analysis, we create 100 “overimputed” datasets using the standard errorsaround the liberalism estimates. We estimate the model on each of the 100 datasets and com-bine inferences about model parameters using standard multiple-imputation formulas (Blackwell,Honaker, and King 2012; Honaker, King, and Blackwell 2011).
42. The estimated responsiveness slope for Southern senators is 0.57, with a standard error of0.29.
43. Strictly speaking, the ideal-point estimates are not comparable across chambers. However,ideal points in both chambers have a roughly standard normal distribution. The standard deviationof the liberalism of state delegations is a bit larger in the House (0.6 versus 0.5 in the Senate).
25
Table 1: Congressional responsiveness to Citizen Liberalism by chamber and region,1936–52. Citizen Liberalism and MC Liberalism are averaged across years within con-gresses. The estimates for Southern state publics exclude black citizens. The modelincludes fixed effects for congress-chamber-region (estimates not shown). Standarderrors are clustered by state. Measurement error is addressed using “overimputation”(Blackwell, Honaker, and King 2012).
Dependent variable:
MC Liberalism
Citizen Liberalism (House) 0.27∗∗
(0.11)
Senate × Citizen Liberalism 0.30(0.33)
Non-South × Citizen Liberalism 0.43∗
(0.25)
Non-South × Senate × Citizen Liberalism −0.44(0.36)
States 48Congresses 9Observations 864R2 0.23Adjusted R2 0.19
∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01
26
conservative “haves” over the more liberal “have-nots.”This analysis of responsiveness to mass policy preferences has also revealed sug-
gestive evidence of differences between the House and Senate. Given that publicopinion must be measured at the level of states rather than districts, however, it isdifficult to know how to interpret heterogeneity across chambers. To further explorethis question, we take advantage of data we do have at the district level: the dis-tribution of household income. In addition to permitting a more direct analysis ofresponsiveness in the House, using income as a proxy for citizens’ economic interestshelps rule out the possibility that citizens’ preferences are derived from the behaviorof their representatives rather than the other way around.
4.4 Responsiveness to Economic Interests
All else equal, policies that make the distribution of economic resources more equaltend to benefit poorer citizens more than richer ones. Consequently, if income isall that voters care about and politicians respond to the median voter, electorateswhose median income is high relative to the average should select representativeswho are less supportive of redistributive policies.(Meltzer and Richard 1981; Roe-mer 1999; Londregan 2006). The expectations provided by a median-voter modelof redistribution thus provide a way to evaluate responsiveness without having tomeasure citizens’ preferences directly.44
On the whole, the New Deal was strongly redistributive, at least by the standardsof American politics. New Deal policies contributed to the decline in economicinequality in the middle of the 20th century, and conflict over these policies helpedrealign political cleavages along more class-based lines (Oestreicher 1998; Sundquist1983). Indeed, Figure 2 indicates that upper-class survey respondents (phone ownersand professionals) were in fact less supportive of New Deal liberalism. Since ourmeasure of MC Liberalism is based on the issue complex at the core of the New
44. Many empirical studies have found little cross-sectional correlation between the extent of re-distribution and income inequality. See, for example, Lindert (2004) on the “Robin Hood paradox,”and Lupu and Pontusson (2011) on the importance of income “skew.” There is substantial evidence,however, for the median-voter model’s comparative statics. Husted and Kenny (1997) find that theelimination of suffrage restrictions in U.S. states led to higher turnout, a poorer median voter, andhigher state spending on social welfare. Fowler (2013) finds similar results in an examination ofcompulsory voting in Australian provinces. In the study most relevant to the one-party South, Fleck(2002) analyzes voting patterns on the 1938 Fair Labor Standards Act, using general-election voterturnout as a proxy for the political influence of low-wage workers. Fleck finds that representativesfrom congressional districts where electoral turnout in 1932 was higher were more likely to supportthe FLSA.
27
Deal, it provides a reasonable indicator of senators’ and representatives’ support foreconomic redistribution.
Of course, New Deal liberalism was not simply about class. It also involved theuse of federal resources to aid underdeveloped areas of the country and, increasingly,conflict over urban–rural divisions and the autonomy of the South’s racial politicaleconomy(Garson 1974; Katznelson, Geiger, and Kryder 1993; Schwartz 1993; Badger2007).45 It is thus important to account for ruralness and the intensity commitmentto white supremacy in our evaluation of the responsiveness of Southern MCs to theeconomic interests of their constituents.
There were three basic sources of variation in median income across Southernelectorates. Most obviously, some areas of the South, particularly those with whiterand more urban population, were wealthier and more developed. Secondly, con-stituencies differed in their degree of income equality; a constituency with relativelyhigh per-capita income might still have relatively low median income if income wasconcentrated at the very top. Thirdly, the size of the electorate relative to the popu-lation differed across constituencies. Holding constant the income of the population,the median voter was richer where turnout was lower due to the strong class bias inSouthern turnout. Our measure of median income in the electorate must take intoaccount these three sources of variation.
Yearly data on the income distribution in congressional districts is not directlyavailable, and we must use a fairly complicated procedure to estimate it (see Ap-pendix A). In brief, the procedure is to, first, use county-level census data to imputethe yearly proportions of households in each year with income below two thresholds:1.3 × national per-capita income (NPCI) and 3.3 × NPCI.46 Second, the analogousconstituency-level proportions are derived by aggregating counties into congressionaldistricts or states. These proportions pin down the entire income distribution un-der the apparently reasonable assumption that income was log-normally distributedwithin constituencies. Finally, the constituency-level income distribution is used toestimate median household income (relative to the nation) in various definitions ofthe electorate. Since we have collected data on the county composition of congres-sional districts in the South only, we confine our analysis to Southern states.
In 1930, the South was substantially poorer than the rest of the nation; themedian household income in the typical congressional district was around 1.2 NPCI.47
45. Compare with Lee and Roemer (2006) on the effects of “policy bundling” between racial andeconomic policies.
46. It is appropriate to normalize income by the national average because we are examiningcongressional representation, and thus the community within which income may be redistributedis the nation, not the constituency.
47. Unless income is very unequally distributed, per-capita income will be lower than median
28
By 1960 some areas, such as Dallas, Texas and Arlington, Virginia, had convergedwith the rest of the nation, achieving median household incomes of over 2 NPCI.48
Most of the South, however, became only a little more prosperous relative to thenation during these years. Throughout this period, the poorest areas of the Southtended to be rural and often highly African American—districts like Mississippi’s 3rd,a black-belt district where the median household income was 0.7 NPCI in 1940.49
The districts with the poorest electorates, however, were rural white districts suchas Arkansas’s 2nd, where the population was poor but largely enfranchised.50
We consider four alternative definitions of the (potential) electorate in Southernconstituencies. The most restrictive definition is the set of voters who participatedin the most recent Democratic congressional primary in the state or district. Thisrequires dropping uncontested primaries from the analysis. A slightly broader def-inition is the set of voters who participated in the most recent presidential generalelection. A still more expansive definition of the electorate is the white public. Fi-nally, the maximal definition includes all residents of a state or district. Althoughnone of these definitions is perfect, the best is probably the white population, whichis less endogenous than the active electorate but captures the potential electoratemore accurately than the total population.51
We estimate the electorate’s median relative income in constituency d and electionyear t using the formula
mdt = exp
[1− Φ
(0.5× edt − µlog
dt
σlogdt
)], (1)
where edt is the size of the electorate as a proportion of the voting-age population,and µlog
dt and σlogdt are, respectively, the estimated mean and standard deviation of
log relative income (assumed normally distributed).52 To evaluate responsiveness to
household income because its denominator is the total number of residents, which is greater thanthe number of households.
48. Arlington’s congressional district, VA-10, was captured by the Republican Party as soon as itwas created in 1952.
49. As it happens, MS-3 was represented in the 1950s by Frank Smith, who was actually quiteliberal on economic issues and moderate (for Mississippi) on racial ones (Badger 2007).
50. AR-2 was represented by Wilbur Mills, longtime chair of Ways and Means and architect ofMedicare and Medicaid (Zelizer 1998).
51. Among other things, primary turnout is directly influenced by the quality of congressionalrepresentation, not least because it can be measured only when primaries are contested. Presidentialturnout avoids this problem but is affected by the competitiveness of presidential elections, whichin this era was correlated with a higher Republican share (and thus a more conservative electorate).
52. This formula also assumes that every member of the electorate (however defined) has higher
29
median voter income, we estimate several models of the following general form:
ydt = β1mdt + β2(mdt × Senated) + f(Y eart, Senated,xdt) + ξs[d] + εdt, (2)
where ydt denotes the average liberalism of MCs representing constituency d in thecongress elected in year t, xdt is a vector of constituency-specific time-varying covari-ates, and ξs[d] is an (optional) fixed effect for the constituency’s state. The quantitiesof interest are β1 (responsiveness in the House), β2 (the difference in responsivenessby chamber), and β1 + β2 (responsiveness in the Senate). We highlight the resultsfrom four specifications:
1. Year FEs: Fixed effects for chamber-year; no constituency controls
2. Linear: Fixed effects for chamber-year;controls for % Black, % Rural, and % Black × % Rural
3. Splines: Restricted cubic spline in Year, separately by chamber;three-way interaction between splines in Year, % Black, and % Rural53
4. State FEs: Same specification as Splines but with fixed effects for State
The combination of four definitions of the electorate (Primary Voters, PresidentialVoters, All Whites, and All Citizens) and four model specifications (Year FEs, Linear,Splines, and State FEs) yields sixteen sets of results. Figure 6 presents the estimatedresponsiveness of Southern senators (β1 +β2) and House members (β1) as well as thedifference between chambers (β2). The standard errors for all of these coefficientsare clustered by member of Congress.54 To account for uncertainty in the incomeimputation, 10 multiply imputed datasets were created and inferences combinedacross them.
Qualitatively, all the specifications tell a similar story. Income generally predictscongressional conservatism in both chambers, with the slope estimates smaller but
income than every non-voter. Notwithstanding the strong class bias in Southern turnout, thisassumption is unlikely to hold perfectly. However, unless the class bias in turnout (conditional onthe overall level of turnout) varies across constituencies, differences in mdt will be monotonicallyrelated to differences in true median voter income.
53. Splines permit flexible non-linear specifications that economize on degrees of freedom (Harrell2001, 20–4).
54. More specifically, the clusters consists of each unique combination of members representing agiven constituency at the same time. For example, each pair of senators who represented the samestate at the same time is assigned a unique ID.
30
Primary Voters Presidential Voters All Whites All Citizens
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Senate − House
House
Senate
Senate − House
House
Senate
Senate − House
House
Senate
Senate − House
House
Senate
Year FE
sLinear
Splines
State F
Es
−0.8 −0.4 0.0 −0.8 −0.4 0.0 −0.8 −0.4 0.0 −0.8 −0.4 0.0Coefficient Estimate for Median Relative Income (90% CI)
Median Income and MC Liberalism
Figure 6:
more precise for primary and presidential voters.55 The magnitude of the pointestimates ranges between 0.01 and 0.56. Substantively, the largest estimates implythat the decrease in MC Liberalism associated with a 1-NPCI increase in voters’median household income is approximately equal to the standard deviation of MCLiberalism within the Southern caucus.56
Though the slope differences between chambers are not not always significant,responsiveness again appears to have been stronger in the Senate than the House.57
This impression is reinforced by the fact that in the specification with state fixedeffects (bottom row), which account for persistent differences between states, theHouse coefficient is reduced to insignificance in all cases. Only when we consider thepresidential electorate is responsiveness approximately equal across chambers.
55. The greater variability of the estimates for whites and citizens maybe due to the “sluggishness”of these variables and their collinearity with demographic controls.
56. MC Liberalism has approximately unit-variance in each chamber. Among SouthernDemocrats, its standard deviation was 0.56 in the House and 0.42 in the Senate.
57. This is probably not a consequence of different ideal-point scales across chambers, since MCLiberalism is actually more variable in the House than the in Senate (where it is usually averagedacross two senators).
31
5 Conclusion
In this paper, I have presented a view of Southern politics quite at odds with theconventional wisdom. In essence, I have argued that MCs from the one-party Southcan be analyzed using the same framework that political scientists use to explaincongressional politics more generally. Like politicians elsewhere in the country, thebehavior of Southern MCs is largely, though by no means exclusively, attributableto their electoral incentives (Mayhew 1974a; cf. Fenno 1973). Moreover, the poten-tial electorate was inclusive enough that Southern MCs responded not only to elitepreferences, but to the preferences of middle- and low-income whites as well.
My thesis would seem to be at odds with the increasingly common practice of la-beling the one-party South an “authoritarian regime” (Farhang and Katznelson 2005;Gibson 2012; Mickey 2014). This is true insofar as authoritarianism is considered topreclude formal opportunities for popular participation and political contestation.58
However, if authoritarianism is conceptualized as encompassing a wide variety ofregimes that fall short of democratic standards, then my argument is perfectly com-patible with a view of the one-party South as an authoritarian regime. Without adoubt, Southern states fell far short of democratic standards, most importantly intheir exclusion of African Americans from political citizenship.
Yet an important premise of my argument is that the political exclusion of South-ern blacks was qualitatively different from suffrage limitations on whites. Thoughpruning “unworthy” whites from the electorate was certainly a goal of many turn-of-the-century disenfranchisers—as it was for contemporaneous proponents of lit-eracy tests and other suffrage qualifications in the North—this goal was never fullyachieved, in part due to the appeal of a countervailing ideology of “herrenvolk democ-racy” for whites. Put simply, even poor whites were considered part of the politicalcommunity in a way that blacks were not.59 Thus, notwithstanding the very real bar-riers to lower-class political participation, with respect to whites the one-party Southlay somewhere on the continuum between autocracy and democracy (and exhibitedsubstantial internal variation on this continuum).
Distinguishing the effects of political exclusion from those of restricted competi-tion is essential to understanding the implications of the one-party South for demo-cratic theory. As the evidence in this paper indicates, lack of partisan competitiondoes not seem to have prevented Southern MCs from responding to the preferences of
58. Mickey (2014) carefully avoids such a definition. Indeed, disagreement and contestation withinthe one-party system are central to his account of the South’s “transition to democracy.”
59. Even politicians who sought black votes could not afford to do so publicly, and so rather didso through emissaries to the black community.
32
the eligible electorate.60 In fact, contrary to the conventional conclusion that lack ofparty competition biased representation in favor of the “haves,” Southern MCs wereactually more liberal than one would expect given the economic policy preferencesof their constituents.
Despite the strength of the evidence for responsiveness, the case for representa-tion without parties must be carefully qualified. Southern MCs did not face regularpartisan competition, but neither did they act in a totally party-less environment.Indeed, the fact that congressional roll-call voting was highly structured by parti-sanship compensated for some of the limitations of completely nonpartisan politics.Southern MCs could not avoid taking clear stances on high-profile, ideologically de-fined decisions, for which they could (with the aid of challengers) be held accountableby the electorate. By contrast, state politics in the South, the focus of Key’s briefagainst one-partyism, was more insulated from the effects of the national two-partysystem.
On a final note, let us consider the implications of these findings for our under-standing of American political development. As Ira Katznelson has argued mostforcefully, Southern MCs pivotal position in national politics gave them outsized rolein shaping the welfare state and political economy of the United States (Katznelson2005, 2013; see also Quadagno 1994; Lieberman 1998; Brown 1999; Alston and Ferrie1999). By and large, these accounts have treated Southern MCs as representing theSouthern elite. They attribute Southern Democrats’ turn against organized labor inthe 1930s and 1940s, for example, to planters and other Southern employers’ con-cern that unions threatened the region’s racial system and labor-repressive politicaleconomy (Katznelson, Geiger, and Kryder 1993; Farhang and Katznelson 2005). AsSchickler and Caughey (2011) observe, however, this explanation ignores the poten-tial impact of the contemporaneous anti-labor turn of public opinion in the South(as well as in the non-South).
My findings in this paper suggest that a fuller account of Southern MCs’ role
60. This study has not considered representation on civil rights issues, on which Southern whiteswere much less divided and were much more clearly national outliers than on economic issues. Butthe basic argument applies at least as strongly in this domain. Dahl (1971, 93–4) rightly notesthat “southern whites were overwhelmingly allegiant” to the South’s system of racial oppression.In a 1942 poll, for example, only 2% of Southern whites reported supporting racially integratedschools, compared to 42% of non-Southern whites (Schuman et al. 1997, 109). White opposition tocivil rights was so overwhelming that Southern MCs almost never deviated from white supremacistorthodoxy (though see Werner 2009). Thus, Miller and Stokes (1963, 55) observe, “the electorate’ssanctions” on civil rights issues were “potential rather than actual. . . . Nevertheless, the fact ofconstituency influence, backed by potential sanctions at the polls, [was] real enough.” They cite asan example Dale Alford’s write-in primary victory over Rep. Brooks Hays (D-AR) after the lattertried to help defuse the 1956 crisis over school integration in Little Rock.
33
in American political development must be rooted in their electoral incentives and,in turn, in the political preferences and behavior of the white public. From thisperspective, the reason that Southern MCs were pivotal in Congress was that theSouthern white public was torn between the benefits and dangers of New Deal liber-alism. Such a framework still leaves ample room for economic elites in the one-partySouth to have exercised disproportionate influence, as indeed they appear to exercisein the contemporary United States (Bartels 2008; Gilens and Page, forthcoming).But by reducing the distance between between the “authoritarian” South and the“democratic” non-South, it enhances our understanding of both.
34
References
Achen, Christopher H. 1978. “Measuring Representation.” American Journal of Po-litical Science 22 (3): 475–510.
Aldrich, John H. 2011. Why Parties? A Second Look. Chicago: University of ChicagoPress.
Aldrich, John H., and John D. Griffin. 2010. “Parties, Elections, and DemocraticPolitics.” Chap. 31 in The Oxford Handbook of American Elections and PoliticalBehavior, edited by Jan E. Leighley, 595–610. New York: Oxford UniversityPress.
Alston, Lee J., and Joseph P. Ferrie. 1999. Southern Paternalism and the AmericanWelfare State: Economic, Politics, and Institutions in the South, 1865–1965.New York: Cambridge University Press.
Alt, James E. 1994. “The Impact of the Voting Rights Act on Black and WhiteVoter Registration in the South.” In Quiet Revolution in the South: The Impactof the Voting Rights Act, 1965–1990, edited by Chandler Davidson and BernardGrofman, 351–377. Princeton, NJ: Princeton UP.
Alvarez, Mike, Jose Antonio Cheibub, Fernando Limongi, and Adam Przeworski.1996. “Classifying Political Regimes.” Studies in Comparative International De-velopment 31 (2): 3–36.
American Political Science Association. 1950. “Toward a More Responsible Two-Party System: A Report of the Committee on Political Parties.” American Po-litical Science Review 44 (3, Part 2, Supplement): 1–96.
Ansolabehere, Stephen, John Mark Hansen, Shigeo Hirano, and James M. SnyderJr. 2010. “More Democracy: The Direct Primary and Competition in U.S. Elec-tions.” Studies in American Political Development 24 (2): 190–205.
Arnold, R. Douglas. 1990. The Logic of Congressional Action. New Haven, CT: YaleUniversity Press.
Badger, Anthony J. 2007. “Whatever Happened to Roosevelt’s New Generation ofSoutherners?” Chap. 4 in New Deal/New South, 58–71. Fayetteville: Universityof Arkansas Press.
35
Bafumi, Joseph, and Michael C. Herron. 2010. “Leapfrog Representation and Extrem-ism: A Study of American Voters and Their Members in Congress.” AmericanPolitical Science Review 104 (3): 519–542.
Bartels, Larry M. 2008. Unequal Democracy: The Political Economy of the NewGilded Age. Princeton, NJ: Princeton University Press.
Bartley, Numan V. 1995. The New South, 1945–1980: The Story of the South’s Mod-ernization. Baton Rouge: Louisiana State University Press.
Bensel, Richard. 1984. Sectionalism and American Political Development: 1880–1980. Madison: University of Wisconsin Press.
Berinsky, Adam J. 2006. “American Public Opinion in the 1930s and 1940s: TheAnalysis of Quota-Controlled Sample Survey Data.” Public Opinion Quarterly70 (4): 499–529.
. 2009. In Time of War: Understanding Public Opinion, From World War IIto Iraq. Chicago: University of Chicago Press.
Berinsky, Adam J., Eleanor Neff Powell, Eric Schickler, and Ian Brett Yohai. 2011.“Revisiting Public Opinion in the 1930s and 1940s.” PS: Political Science &Politics 44 (3): 515–520.
Besley, Timothy, and Stephen Coate. 1997. “An Economic Model of RepresentativeDemocracy.” Quarterly Journal of Economics 112 (1): 85–114.
Black, Earl, and Merle Black. 1987. Politics and Society in the South. Cambridge,MA: Harvard University Press.
. 2002. The Rise of Southern Republicans. Cambridge, MA: Belknap Press.
Blackwell, Matthew, James Honaker, and Gary King. 2012. “Multiple Overimpu-tation: A Unified Approach to Measurement Error and Missing Data.” http:
//gking.harvard.edu/files/measure.pdf.
Bloch Rubin, Ruth. 2014. “Securing the Solid South: Intraparty Organization andthe Development of the Southern Caucus and Southern Delegations.” Paperpresented at the 2014 Congress and History Conference, College Park, Maryland.
Brady, David W., and Charles S. Bullock III. 1980. “Is There a Conservative Coalitionin the House?” Journal of Politics 42 (2): 549–559.
36
Brattain, Michelle. 1997. “Making Friends and Enemies: Textile Workers and Politi-cal Action in Post-World War II Georgia.” Journal of Southern History 63 (1):91–138.
Brooks, Jennifer E. 2000. “Winning the Peace: Georgia Veterans and the Struggle toDefine the Political Legacy of World War II.” Journal of Southern History 66(3): 563–604.
Brown, Michael K. 1999. Race, Money, and the American Welfare State. Ithaca, NY:Cornell University Press.
Bueno de Mesquita, Bruce, Alastair Smith, Randolph M. Siverson, and James D.Morrow. 2003. The Logic of Political Survival. Cambridge, MA: MIT Press.
Bullock, Charles S., III, and Mark J. Rozell, eds. 2012. The Oxford Handbook ofSouthern Politics. New York: Oxford UP.
Carleton, William G. 1951. “The Southern Politician—1900 and 1950.” Journal ofPolitics 13, no. 2 (May): 215–231.
Caughey, Devin, and Eric Schickler. 2014. “Structure and Change in CongressionalIdeology: NOMINATE and Its Alternatives.” Paper presented at the 2014 Congressand History Conference, University of Maryland, College Park, June 11, 2014.
Caughey, Devin, and Mallory Wang. 2014. “Bayesian Population Interpolation andLasso-Based Target Selection in Survey Weighting.” Paper presented at theSummer Meeting of the Society for Political Methodology, University of Georgia,Athens, GA, July 24–26, 2014.
Caughey, Devin, and Christopher Warshaw. 2013. “Dynamic Estimation of LatentOpinion from Sparse Survey Data Using a Group-Level IRT Model.” Paperpresented at the Summer Meeting of the Society for Political Methodology, Uni-versity of Virginia, Charlottesville, VA, July 18–20, 2013.
Christian, Ralph J. 1976. “The Folger-Chatham Congressional Primary of 1946.”North Carolina Historical Review 53 (1): 25–54.
Clausen, Aage R. 1973. How Congressmen Decide: A Policy Focus. New York: St.Martin’s Press.
Clinton, Joshua D., and Simon Jackman. 2009. “To Simulate or NOMINATE?” Leg-islative Studies Quarterly 34, no. 4 (November): 593–621.
37
Cobb, James C. 2010. The South and America since World War II. New York: OxfordUP.
Coode, Thomas H. 1971. “Bryant Thomas Castellow of Georgia.” Georgia Advocate8 (Fall): 16–20.
Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven, CT:Yale University Press.
Dangerfield, Royden J., and Richard H. Flynn. 1936. “Voter Motivation in the 1936Oklahoma Democratic Primary.” Southwestern Social Science Quarterly 17 (2):97–105.
Domhoff, G. William. 1996. State Autonomy or Class Dominance? New York: Aldine.
Domhoff, G. William, and Michael J. Webber. 2011. Class and Power in the NewDeal: Corporate Moderates, Southern Democrats, and the Liberal-Labor Coali-tion. Stanford, CA: Stanford University Press.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper.
Dunn, Susan. 2010. Roosevelt’s Purge: How FDR Fought to Change the DemocraticParty. New York: Cambridge University Press.
Erskine, Hazel Gaudet. 1964. “The Polls: Some Gauges of Conservatism.” PublicOpinion Quarterly 28 (1): 154–168.
Ewing, Cortez A. M. 1953. Primary Elections in the South: A Study in UnipartyPolitics. Norman: University of Oklahoma Press.
Farhang, Sean, and Ira Katznelson. 2005. “The Southern Imposition: Congress andLabor in the New Deal and Fair Deal.” Studies in American Political Develop-ment 19 (1): 1–30.
Fearon, James D. 1999. “Electoral Accountability and the Control of Politicians:Selecting Good Types versus Sanctioning Poor Performance.” In Democracy,Accountability, and Representation, edited by Adam Przeworski, Susan CarolStokes, and Bernard Manin. New York: Cambridge University Press.
Fenno, Richard F., Jr. 1973. Congressmen in Committees. Boston: Little, Brown.
Ferejohn, John. 1986. “Incumbent Performance and Electoral Control.” Public Choice50:5–25.
38
Finley, Keith M. 2008. Delaying the Dream: Southern Senators and the Fight againstCivil Rights, 1938–1965. Baton Rouge: Louisiana State University Press.
Fleck, Robert K. 2002. “Democratic Opposition to the Fair Labor Standards Act of1938.” Journal of Economic History 62 (1): 25–54.
Flinn, Thomas A., and Harold L. Wolman. 1966. “Consistency and Roll Call Voting:The Case of Southern Democratic Congressmen.” Midwest Journal of PoliticalScience 10 (2): 192–199.
Fowler, Anthony. 2013. “Electoral And Policy Consequences Of Voter Turnout: Ev-idence From Compulsory Voting In Australia.” Quarterly Journal of PoliticalScience 8 (2): 159–182.
Gailmard, Sean, and Jeffery A. Jenkins. 2009. “Agency Problems, the 17th Amend-ment, and Representation in the Senate.” American Journal of Political Science53 (2): 324–342.
Garson, Robert A. 1974. The Democratic Party and the Politics of Sectionalism,1941–1948. Baton Rouge: Louisiana State University Press.
Gelman, Andrew, and Thomas C. Little. 1997. “Poststratification Into Many Cate-gories Using Hierarchical Logistic Regression.” Survey Methodology 23 (2): 127–135.
Gibson, Edward L. 2012. Boundary Control: Subnational Authoritarianism in FederalDemocracies. New York: Cambridge UP.
Gilens, Martin, and Benjamin I. Page. Forthcoming. “Testing Theories of AmericanPolitics: Elites, Interest Groups, and Average Citizens.” Perspectives on Politics.
Grynaviski, Jeffrey D. 2004. “The Impact of Electoral Rules on Factional Competi-tion in the Democratic South, 1919–48.” Party Politics 10 (5): 499–519.
Hardeman, D. B., and Donald C. Bacon. 1987. Rayburn: A Biography. Austin: TexasMonthly Press.
Harrell, Frank E., Jr. 2001. Regression Modeling Strategies: With Applications toLinear Models, Logistic Regression, and Survival Analysis. New York: Springer-Verlag.
Hofstadter, Richard. 1949. “From Calhoun to the Dixiecrats.” Social Research 16(1): 135–150.
39
Honaker, James, Gary King, and Matthew Blackwell. 2011. “Amelia II: A Programfor Missing Data.” Journal of Statistical Software 45 (7): 1–47. http://www.jstatsoft.org/v45/i07/.
Husted, Thomas A., and Lawrence W. Kenny. 1997. “The Effect of the Expansion ofthe Voting Franchise on the Size of Government.” Journal of Political Economy105 (1): 54–82.
Katz, Richard S. 2006. “Party in Democratic Theory.” In Handbook of Party Politics,34–46. Thousand Oaks, CA: Sage.
Katznelson, Ira. 2005. When Affirmative Action Was White: An Untold History ofRacial Inequality in Twentieth-Century America. New York: W. W. Norton.
. 2013. Fear Itself: The New Deal and the Origins of Our Time. New York:Liveright.
Katznelson, Ira, Kim Geiger, and Daniel Kryder. 1993. “Limiting Liberalism: TheSouthern Veto in Congress, 1933–1950.” Political Science Quarterly 108 (2):283–306.
Katznelson, Ira, and Quinn Mulroy. 2012. “Was the South Pivotal? Situated Parti-sanship and Policy Coalitions during the New Deal and Fair Deal.” Journal ofPolitics 74 (2): 604–620.
Key, V. O., Jr. 1949. Southern Politics in State and Nation. New York: Knopf.
. 1964. “Congressional Nominations.” In Politics, Parties & Pressure Groups,434–454. New York: Crowell.
Kousser, J. Morgan. 1974. The Shaping of Southern Politics: Suffrage Restrictionand the Establishment of the One-Party South. New Haven, CT: Yale UniversityPress.
. 1984. “Origins of the Run-Off Primary.” The Black Scholar 15 (5): 23–26.
Lee, Woojin, and John E. Roemer. 2006. “Racism and Redistribution in the UnitedStates: A Solution to the Problem of American Exceptionalism.” Journal ofPublic Economics 90 (6-7): 1027–1052.
Leip, David. 2013. Dave Leip’s Atlas of U.S. Presidential Elections. http://www.uselectionatlas.org.
Lenz, Gabriel. 2012. Follow the Leader? How Voters Respond to Politicians’ Perfor-mance and Policies. Chicago: University of Chicago Press.
40
Levitsky, Steven, and Maxwell A. Cameron. 2003. “Democracy Without Parties? Po-litical Parties and Regime Change in Fujimori’s Peru.” Latin American Politicsand Society 45 (3): 1–33.
Lieberman, Robert C. 1998. Shifting the Color Line: Race and the American WelfareState. Cambridge, MA: Harvard University Press.
Lindert, Peter H. 2004. Growing Public: Social Spending and Economic Growth sincethe Eighteenth Century. New York: Cambridge University Press.
Lipset, Seymour Martin. 1996. “What Are Parties For? Review of Why Parties? TheOrigin and Transformation of Party Politics in America, by John Aldrich.”Journal of Democracy 7 (1): 169–175.
Londregan, John. 2006. “Political Income Distribution.” Chap. 5 in The OxfordHandbook of Political Economy, edited by Barry R. Weingast and Donald A.Wittman, 84–101. New York: Oxford UP.
Lupu, Noam, and Jonas Pontusson. 2011. “The Structure of Inequality and the Pol-itics of Redistribution.” American Political Science Review 105 (2): 316–336.
Manion, Melanie. 2014. “‘Good Types’ in Authoritarian Elections: The SelectoralConnection in Chinese Local Congresses.” Comparative Political Studies 47:1–33.
Manza, Jeff. 2000. “Political Sociological Models of the U.S. New Deal.” AnnualReview of Sociology 26:297–322.
Martin, Andrew D., and Kevin M. Quinn. 2002. “Dynamic Ideal Point Estimation viaMarkov Chain Monte Carlo for the U.S. Supreme Court, 1953–1999.” PoliticalAnalysis 10 (2): 134–153.
Martin, Andrew D., Kevin M. Quinn, and Jong Hee Park. 2011. “MCMCpack: MarkovChain Monte Carlo in R.” Journal of Statistical Software 42 (9): 1–21.
Matsusaka, John G. 2001. “Problems with a Methodology Used to Evaluate the VoterInitiative.” Journal of Politics 63 (4): 1250–1256.
Mayhew, David R. 1966. Party Loyalty Among Congressmen: The Difference BetweenDemocrats and Republicans, 1947–1962. Cambridge, MA: Harvard UniversityPress.
. 1974a. Congress: The Electoral Connection. New Haven, CT: Yale UP.
41
Mayhew, David R. 1974b. “Congressional Elections: The Case of the VanishingMarginals.” Polity 6 (3): 295–317.
McDonald, Michael P., and Samuel L. Popkin. 2001. “The Myth of the VanishingVoter.” American Political Science Review 95 (4): 963–974.
Meltzer, Allan H., and Scott F. Richard. 1981. “A Rational Theory of the Size ofGovernment.” Journal of Political Economy 89 (5): 914–927.
Mickey, Robert W. 2008. “The Beginning of the End for Authoritarian Rule in Amer-ica: Smith v. Allwright and the Abolition of the White Primary in the DeepSouth, 1944–1948.” Studies in American Political Development 22 (2): 143–182.
. 2014. Paths Out of Dixie: The Democratization of Authoritarian Enclavesin America’s Deep South. Princeton, NJ: Princeton UP.
Mill, John Stuart. 1867. Considerations on Representative Government. People’s.London: Longmans, Green.
Miller, Warren E., and Donald E. Stokes. 1963. “Constituency Influence in Congress.”American Political Science Review 57 (1): 45–56.
Morgan, Chester M. 1985. Redneck Liberal: Theodore G. Bilbo and the New Deal.London: Louisiana State University Press.
Nixon, H. C. 1946. “Politics of the Hills.” Journal of Politics 8 (2): 123–133.
Norrell, Robert J. 1991. “Labor at the Ballot Box: Alabama Politics from the NewDeal to the Dixiecrat Movement.” Journal of Southern History 57 (2): 201–234.
Oestreicher, Richard. 1998. “The Rules of the Game: Class Politics in Twentieth-Century America.” In Organized Labor and American Politics, 1894-1994: TheLabor-Liberal Alliance, edited by Kevin Boyle. Albany: SUNY Press.
Park, David K., Andrew Gelman, and Joseph Bafumi. 2004. “Bayesian MultilevelEstimation with Poststratification: State-Level Estimates from National Polls.”Political Analysis 12 (4): 375–385.
Patterson, James T. 1967. Congressional Conservatism and the New Deal: The Growthof the Conservative Coalition in Congress, 1933–1939. Lexington: University ofKentucky Press.
Pitkin, Hanna Fenichel. 1967. The Concept of Representation. Berkeley: Universityof California Press.
42
Poole, Keith T., and Howard Rosenthal. 1998. HCODES.TXT. Text file downloadedfrom http://voteview.com/page2c.htm.
. 2001. Houses 1 - 106 Outcome Coordinates and Issue Codes. ftp://voteview.com/h01106xx.dat. Last updated 3 July 2001. Accessed 7 July 2011.
. 2007. Ideology & Congress. New Brunswick, NJ: Transaction Publishers.
Quadagno, Jill. 1994. The Color of Welfare. New York: Columbia University Press.
Roemer, John E. 1999. “The Democratic Political Economy of Progressive Taxation.”Econometrica 67 (1): 1–19.
Rosenfarb, Joseph. 1944. “Labor’s Role in the Election.” Public Opinion Quarterly8 (3): 376.
Sanders, Elizabeth. 1999. Roots of Reform: Farmers, Workers, and the AmericanState, 1877–1917. Chicago: University of Chicago Press.
Schattschneider, E. E. 1942. Party Government. Westport, CT: Greenwood Press.
Schickler, Eric, and Devin Caughey. 2011. “Public Opinion, Organized Labor, andthe Limits of New Deal Liberalism, 1936–1945.” Studies in American PoliticalDevelopment 25 (2): 1–28.
Schickler, Eric, and Kathryn Pearson. 2009. “Agenda Control, Majority Party Power,and the House Committee on Rules, 1937–52.” Legislative Studies Quarterly 34(4): 455–491.
Schuman, Howard, Charlotte Steeh, Lawrence Bobo, and Maria Krysan. 1997. RacialAttitudes in America: Trends and Interpretations. Revised. Harvard UniversityPress.
Schwartz, Jordan A. 1993. The New Dealers: Power Politics in the Age of Roosevelt.New York: Vintage Books.
Sen, Amatrya. 1999. Development as Freedom. New York: Knopf.
Shafer, Byron E., and Richard Johnston. 2006. The End of Southern Exceptional-ism: Class, Race, and Partisan Change in the Postwar South. Cambridge, MA:Harvard University Press.
Snyder, James M., Jr., and Michael M. Ting. 2002. “An Informational Rationale forPolitical Parties.” American Journal of Political Science 46 (1): 90–110.
43
Snyder, James M., Jr., and Michael M. Ting. 2003. “Roll Calls, Party Labels, andElections.” Political Analysis 11 (4): 419–444.
Stimson, James A., Michael B. MacKuen, and Robert S. Erikson. 1995. “DynamicRepresentation.” American Political Science Review 89 (3): 543–565.
Sundquist, James L. 1983. Dynamics of the Party System: Alignment and Realign-ment of Political Parties in the United States. Revised. Washington, DC: Brook-ings.
Swain, Martha H. 2001. “Hubert D. Stephens: Mississippi’s ‘Quiet Man’ in the Sen-ate, 1923–1935.” Journal of Mississippi History 63 (4).
Sweeney, James R. 1999. “‘Sheep without a Shepherd’: The New Deal Faction in theVirginia Democratic Party.” Presidential Studies Quarterly 29 (2): 438–458.
Valelly, Richard M. 2004. The Two Reconstructions: The Struggle for Black Enfran-chisement. Chicago: University of Chicago Press.
Weiss, Stuart L. 1971. “Maury Maverick and the Liberal Bloc.” The Journal ofAmerican History 57 (4): 880–895.
Weissberg, Robert. 1978. “Collective vs. Dyadic Representation in Congress.” Amer-ican Political Science Review 72 (2): 535–547.
Werner, Timothy. 2009. “Congressmen of the Silent South: The Persistence of South-ern Racial Liberals, 1949–1964.” Journal of Politics 71 (1): 70–81.
Woodward, C. Vann. 1951. Origins of the New South, 1877–1913. Baton Rouge:Louisiana State University Press.
Wright, Gerald C., and Brian F. Schaffner. 2002. “The Influence of Party: Evidencefrom the State Legislatures.” American Political Science Review 96 (2): 367–379.
Zelizer, Julian E. 1998. Taxing America: Wilbur D. Mills, Congress, and the State,1945–1975. New York: Cambridge University Press.
44
A Estimation and Imputation of Income
The analysis of Southern MCs’ responsiveness to the economic interests of theirelectorates relies on several proxies for the median income in the electorate. Thissection describes how these proxies were estimated and multiply imputed.
Like all demographic characteristics of congressional districts in the dataset forthe paper, districts’ estimated income distribution must be derived from data atthe county level. Before 1962, all but a handful of Southern congressional districtswere composed of whole counties.61 Based on information contained in Martis et al.(1982), I assigned each county in the 17-state South to its House district in eachcongressional election 1930–62. Using this mapping, it is possible to calculate thedistrict-level value of any quantity that is a function of data available at the countylevel. For example, to calculate the proportion of district residents who are white,I take the weighted average of the proportion white in the counties that composedthe district, where the weights are given by the number of residents of each county.
The calculation of the districts’ income distribution is not so straightforward, forseveral reasons. First, county-level income data are not available before 1950, whenthe U.S. Census began collecting county-level data on median household income(which is missing for some counties) as well as on the proportions of households withyearly income below $2,000 and $5,000. Second, estimating the income of the medianvoter requires the whole income distribution, not just a single statistic such as themean. We deal with these problems with a combination of multiple imputation andassumptions about the shape of the income distribution.
Let’s start with 1950, when income data are observed. We can derive each dis-trict’s proportions of household with income below $2,000 and $5,000 by taking aweighted average of the corresponding quantities for counties in that district.62 Sincewe ultimately care not about absolute income but income relative to the nationalaverage, it is helpful to divide $2,000 and $5,000 by nation per capita income in1950, which was $1,516.63 Let p1.3c,1950 and p3.3c,1950 respectively denote the proportionof county c’s households with relative income below 1.3 ($2, 000/$1, 516) and 3.3($5, 000/$1, 516) in 1950. Denote the analogous district-level quantities as p1.3d,1950
and p3.3d,1950, which are population-weighted averages of the counties that composedistrict d. Together, p1.3d,1950 and p3.3d,1950 provide information on how rich district d is
61. In the 13-state South, the exceptions are districts in New Orleans, Houston, and Miami. Splitcounties were more common in the non-South.
62. Strictly speaking the counties should be weighted by their number of households, but I insteaduse number of residents, which is almost exactly proportional to the number of households.
63. We divide by per capita national income rather than per household because only the formeris available in every year.
45
as well as how unequal it is.The distribution of income is usually considered to be log-normal (i.e., the natural
log of income is normally distributed), a regularity known as “Gibrat’s law.”AssumingGibrat’s law held within Southern congressional districts, p1.3d,1950 and p3.3d,1950 are suffi-cient to pin down the entire income distribution.64 Specifically, the mean of (normallydistributed) log relative income in district d is
µlogd,1950 =
log(1.3)× Φ(p3.3d,1950)− log(3.3)× Φ(p1.3d,1950)
Φ(p3.3d,1950)− Φ(p1.3d,1950), (3)
and its standard deviation is
σlogd,1950 =
log(3.3)− log(1.3)
Φ(p3.3d,1950)− Φ(p1.3d,1950), (4)
where Φ indicates the standard normal CDF.Given µlog
d,1950 and σlogd,1950 and the assumption that log relative income is normally
distributed within districts, we can derive the median income in any subset of theincome distribution. To estimate the median log income the electorate, we mustassume that everyone in the electorate has a higher income than everyone outsidethe electorate—i.e., the class bias is extreme as possible, given the size of the elec-torate. Under this assumption, the median log income (=log median income) in theelectorate is
log(md,1950) = 1− Φ
(0.5× ed,1950 − µlog
d,1950
σlogd,1950
), (5)
where md,1950 is 1950 median income in d’s electorate and ed,1950 is the size the 1950electorate as a proportion of the population.
To estimate median voter income in years other than 1950, we must (multiply)impute the county-level quantities p1.3c,1950 and p3.3c,1950. This requires county-level dataon quantities that are strong predictors of county-level income in 1950 and are avail-able in the 1930, 1940, and 1960 censuses as well. The following variables fit thesecriteria:
• Population
• Proportion Urban
• Proportion White
64. Battistin, Blundell, and Lewbel (2009) provide evidence that Gibrat’s law is approximatelytrue in the United States, and almost exactly so for the top 75% of the income distribution.
46
• State Per Capita Income
• Proportion Enrolled in School (among 14–17 year olds)
Since we wish to predict relative income, these variables must be demeaned withinyears so as to eliminate over-time changes common to all units.
The imputation model specification (with demeaned variables marked with anasterisk),
pxct ∼ N(Population∗ct + Proportion Urban∗
ct + (Proportion Urban∗ct)
2
+ Proportion White∗ct + (Proportion White∗
ct)2
+ Proportion Urban∗ct × Proportion White∗
ct
+ State Per Capita Income∗ct + log(State Per Capita Income∗
ct)
+ Proportion Enrolled in School∗ct), x ∈ {1.3, 3.3} (6)
explains 69% of the variance of p1.3c,1950 and 51% of the variance of p3.3c,1950. To reflectour belief that a county’s relative income distribution is stable over time, we addan informative Gaussian prior for pxct centered on county c’s value of pxc,1950. The
standard deviation of the prior, 0.02×√|1950− t|, implies a belief that the typical
yearly change in pxct was 0.02.The imputation model was estimated using the R package Amelia (Honaker,
King, and Blackwell 2011). Ten imputed county-level datasets were created fort ∈ {1930, 1940, 1960}. In each imputed dataset, values of p1.3ct and p3.3ct for non-census years were linearly interpolated from the census estimates. The estimates foreach year and each imputed dataset were used to calculate the district-level incomedistribution using the same method described above. This yielded ten imputeddistrict-level datasets with estimates of µlog
dt , σlogdt , and mdt.
B Supplementary Figures
47
Ideological Distribution in House (1931−1962)
Congress (Election Year)
Eco
nom
ic L
iber
alis
m
−2
−1
0
1
2
MM M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
MM
M M M M M
M MM
M M M M M
M M
M
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
MM M
MM
M M
M
MM M
MM
M M
M M M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
M
M M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M MM
MM M
M
MM M
MM
M M
M
MM M M
M MM
M M M M M
M MM
M M M MM
M M
M
MM M
MM
M M
M
MM M
M
M MM
M M M M M
M MM
M M M M M
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV VV V
V
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V
V VV
VV
V
V VV V
V
VV
V
V V
VV
VV
VV V VV
V
VV
VV V V
VV
VV
V
V VV V
V
VV
V
V VV
V
VV
VV V VV
V
VV
VV V VV V
VV
V
V VV V
V
VV
V
V VV V
VV
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
V
VV
V VV V
V
VV
V
V VV V
V
V
VV V VV
V
VV
VV V VV
V
VV
V
V VV V
V
VV
V
V VV V
V
VV
V V VV
V
VV
VV V VV
V
VV
V
72(1930)
73(1932)
74(1934)
75(1936)
76(1938)
77(1940)
78(1942)
79(1944)
80(1946)
81(1948)
82(1950)
83(1952)
84(1954)
85(1956)
86(1958)
87(1960)
Party Group
ND
NR
SD
48
Non
−S
outh
ern
Whi
te D
emoc
rat
Sou
ther
n W
hite
Dem
ocra
tN
on−
Sou
ther
n B
lack
Dem
ocra
tS
outh
ern
Bla
ck D
emoc
rat
Non
−S
outh
ern
Whi
te O
ther
/Non
−V
oter
/Mis
sing
Sou
ther
n W
hite
Oth
er/N
on−
Vot
er/M
issi
ngN
on−
Sou
ther
n B
lack
Oth
er/N
on−
Vot
er/M
issi
ngS
outh
ern
Bla
ck O
ther
/Non
−V
oter
/Mis
sing
Non
−S
outh
ern
Whi
te R
epub
lican
Sou
ther
n W
hite
Rep
ublic
anN
on−
Sou
ther
n B
lack
Rep
ublic
anS
outh
ern
Bla
ck R
epub
lican
−2
−1012
−2
−1012
−2
−1012
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
'36
'40
'44
'48
'52
Year
Estimated Average Liberalism in Group
Fig
ure
7:L
iber
alis
mof
diff
eren
tvo
ter
typ
es,
by
race
and
regi
on
49
References
Battistin, Erich, Richard Blundell, and Arthur Lewbel. 2009. “Why Is ConsumptionMore Log Normal than Income? Gibrat’s Law Revisited.” Journal of PoliticalEconomy 117 (6): 1140–1154.
Honaker, James, Gary King, and Matthew Blackwell. 2011. “Amelia II: A Programfor Missing Data.” Journal of Statistical Software 45 (7): 1–47. http://www.jstatsoft.org/v45/i07/.
Martis, Kenneth C., Clifford Lee Lord, Ruth Anderson Rowles, and Historical RecordsSurvey (New York N.Y.) 1982. The Historical Atlas of United States Congres-sional Districts, 1789–1983. New York: Free Press.
50