naaj vol 5 no 3 july 2001, macroeconomic aspects of private

MACROECONOMIC ASPECTS OF PRIVATE

RETIREMENT PROGRAMSKrzysztof M. Ostaszewski*

ABSTRACT

The decline in importance of private defined benefit plans in relation to defined contribution plansin the United States is a major issue of interest to pension actuaries. This decline has been attributedto numerous factors: costs of government regulation, societal and cultural changes, changedemployer attitudes, and employees’ lack of understanding of defined benefit plans. It has alsocaused some observers to proclaim the end of private defined benefit plans. This paper analyzespossible macroeconomic factors contributing to the crisis of defined benefit plans and proposes analternative hypothesis for the cause of the crisis: the decline of the relative attractiveness of definedbenefit plans in relation to defined contribution plans when these are viewed as investments, thatis, as securities in capital markets.

1. INTRODUCTION: ARE DEFINED BENEFIT

PENSION PLANS AN ENDANGERED SPECIES?For more than two decades, traditional definedbenefit (DB) plans in the United States have beendeclining in relative importance. As indicated bythe U.S. Department of Labor (2000), during theperiod 1977–96 the number of DB plans in theUnited States fell at an average annual rate of3.19%, while the number of defined contribution(DC) plans grew in the same period at an annualrate of 4.14%. During the period 1975–96, totalDB plan assets grew at a nominal annual rate of10.23%, while total DC plan assets grew at a nom-inal annual rate of 14.83%. In 1975 total contri-butions to DB plans were $24.2 billion, nearlytwice the $12.8 billion total contributions to DCplans in the same year. In 1996 total contribu-tions to DB plans amounted to $35.8 billion,nearly one-fourth of the $133.7 billion total 1996contributions to DC plans. Note that contribu-tions to DB plans were in a steady decline innominal dollars from 1984 through 1990, andincreases in such contributions have occurredonly in the period 1991–93, followed by a re-sumed decline.

Finally, the total amount of benefits paid by DBplans in 1975 was more than double that for DCplans ($12.9 billion versus $6.2 billion), but in1996 benefits paid by DC plans ($116.5 billion)exceeded those paid by DB plans ($96.9 billion)for the second year in a row (the first time DCbenefits paid ever exceeded DB benefits paid wasin 1995). In 1996, total assets of DB plans stillconstituted the majority (50.6%) of pension planassets, but the trends indicated an imminent tri-umph of DC plans in terms of asset size. Of par-ticular significance was the sizable increase in theassets of the 401(k) variety of DC plans (Poterba,Venti, and Wise 1994).

Figure 1, based on U.S. Department of Labor(2000) data, shows the relative share of pensionplan assets held by DB plans, the portion of allnew pension plan contributions that went into DBplans, the share of all pension benefits paid by DBplans, and the number of DB plan participants asa portion of the number of all pension plan par-ticipants. The figure clearly illustrates a strongshift toward DC plans in the last 22 years.

2. WHY IS THIS HAPPENING?In a debate at a March 1988 meeting of the Ca-nadian Institute of Actuaries moderated by ShirazBharmal (1988) the following resolution was pro-posed: “Defined benefit plans are not well suited

* Krzysztof M. Ostaszewski, Ph.D., F.S.A., M.A.A.A., is Actuarial Pro-gram Director at Illinois State University, Normal, IL 61790-4520,e-mail: [email protected].

52

to tomorrow’s work force and should be replacedby defined contributions plans.” Two of the par-ticipants in the debate, Tom Holmes and DanielMcCaw, argued in favor of the resolution, present-ing a list of factors causing the demise of DBplans. The list was indeed inclusive of what iscommonly believed to be the causes of the shift toDC plans. One should note, however, that, incontrast with the United States, DB plans havenot declined significantly in relation to DC plansin Canada.

The factors presented in the debate led byBharmal (1988) are exactly the ones commonlybrought forth to explain changes in pension struc-ture in the United States and for that reason arepresented here. (Let me stress that I do not en-dorse any of these theories at this point, butmerely present them.)

2.1 Change in the Relationship betweenEmployer and EmployeesToday’s workers are much less likely to retirefrom the same firm in which they began theircareers, while firms are much less likely to viewemployees as lifetime workers. The result is thatemployees need direct ownership of their retire-ment accounts and full portability.

Furthermore, changes in culture, technology,and education have caused employees to be moreself-directed and independent, making DC plansmore desirable. This trend expresses itself both inthe greater number of self-employed persons andin the severing of the bond between employersand employees; both groups are expressing more

direct concern with their economic self-interest.DC plans offer an explicit and concrete measureof wealth, thus better matching the psychology ofthe new workplace environment. Also, DC planstend to be more easily portable. This idea will becalled the New Economy Theory.

2.2 New Tax Laws and FundingRegulations Have Decreased theAttractiveness of DB PlansThe passing of the Employee Retirement and In-come Security Act (ERISA) of 1974 began theprocess of imposition of stricter legal, funding,and solvency requirements in the United States,including establishment of the Pension BenefitGuaranty Corporation (Marcus 1987). Similar re-quirements in Canada were put in place by thePension Benefit Acts. Those stricter funding re-quirements also were accompanied by limitingaccess to surplus by employers. Accounting stan-dards introduced by the Financial AccountingStatement Board in the United States and theCanadian Institute of Chartered Accountants inCanada reduced the flexibility of employers inaccounting valuation of plans and brought uponthem the necessity to perform both accountingand statutory valuations for DB plans, thus im-posing additional costs.

The rules are made additionally complicatedand costly by tax laws concerning deductibility ofemployer contributions to pension plans. Finally,both Canada and the United States impose limi-tations on the tax deductibility of contributions toother retirement accounts for workers participat-ing in a DB plan, which may reduce perceivedattractiveness of DB plans, even though the re-strictions apply equally to DC plans. In addition,Daniel McCaw of William M. Mercer Inc. (Bhar-mal 1988) points out that in Canada legislatorsimposed mandatory inflation indexing on DBplans while making it voluntary for DC plans.McCaw concludes that, in his opinion, legislatorsin Canada are moving often unwilling employerstoward DC plans.

The explanation of the historical shift from DBplans to DC plans as caused by government in-volvement appears to be the most popular amongactuarial professionals. One of the main purposesof the present paper is to challenge this explana-tion (even though one must admit some of its

Figure 1Shares of Pensions in DB Plans

53MACROECONOMIC ASPECTS OF PRIVATE RETIREMENT PROGRAMS

merits). For the sake of simplicity, this set ofideas will be called the Excessive RegulationTheory.

2.3 Distribution of Risk Between Employerand EmployeeThe last 20 years have been marked by unprece-dented volatility in financial markets. This hascaused the cost of funding retirement benefits tobe less predictable. One could argue that, undersuch circumstances, employees should value pro-tection from uncertainty and seek refuge in DBplans. Indeed, such an argument is put forth inthe Canadian Institute of Actuaries’ debate mod-erated by Bharmal (1988) by Patrick Longhurst ofWatson Wyatt Worldwide, who stressed that DCplan participants face the following flaws in theirplans:

● Extreme volatility of benefit levels dependingon investment performance.

● Significant impact of the choice of investmentstrategy.

● Inflexibility for employees hired in midcareer,who cannot be offered pension benefits beyondcontribution limits.

● Difficulty in dealing with early retirement ifsufficient assets have not been accumulated.

If, however, employers are fully aware of the riskdistribution between DB and DC plans (while em-ployees are less than fully informed), have be-come more risk-averse, and have full control ofthe type of the pension plan adopted, then a shiftto DC plans would be expected. This will be calledthe Risk-Averse Employers Theory.

As defined by Longhurst (Bharmal 1988), theunit benefit equals the pension as a percentage ofsalary at retirement, divided by the years of ser-vice, assuming a portfolio invested 50% in bondsand 50% in equities. The unit benefit is a measureof how much of a pension an employee can earnfor each year of service by investing in the bench-mark portfolio. Some historical values of this in-dex for selected years are shown in Table 1. Thetable shows that funding of retirement has be-come cheaper in this period (i.e., the values in thetable are, generally speaking, trending upward)but also somewhat unpredictable. Given thatthere was a massive movement to DC plans dur-ing the period, it can be argued that either em-

ployers were uncomfortable with increased in-vestment risk and preferred to shift it toemployees, or that employees wanted to benefitfrom high returns available in capital markets.

This paper will return to the question of em-ployees’ preferences, but note that the data raisessome doubt about the idea that employers havebeen abandoning DB plans because of high ex-penses. Higher regulatory and legal expenses didindeed happen, but they coincided with higherreturns available in the markets. In other words,pensions may have become more expensive be-cause of having to pay actuaries and lawyersmore, but they also have become cheaper be-cause of higher returns. This makes the ExcessiveRegulation Theory somewhat less credible.

What is suspiciously missing from all of theabove theories of the relative decline of DB planimportance is any economic motivation on thepart of employees. Arguably, the New EconomyTheory refers to the greater economic indepen-dence of individual workers (independence thatmay be wanted or unwanted, depending on theperspective), but it contains no reference to therational self-interest of workers as economic de-cision makers. This absence in the actuarial lit-erature is quite perplexing, and this paper willendeavor to address it.

3. ECONOMIC PERSPECTIVE ON PENSIONS

From the perspective of financial economics, par-ticipation in a pension plan is just another capitalasset, a security similar to a Treasury bill, a shareof stock, or a convertible bond. As in the case ofany other security, plan participants give up to-day’s cash flows (i.e., today’s consumption) inreturn for future cash flows (i.e., future consump-tion). The major difference between DB pension

Table 1Pension Unit Benefit

Year Unit Benefit

1977 0.01111979 0.01221981 0.01831983 0.01391985 0.01441987 0.0173

54 NORTH AMERICAN ACTUARIAL JOURNAL, VOLUME 5, NUMBER 3

plan participation and other securities is that thederivation of future cash flows of a DB plan par-ticipation is based on wages, not open marketcapital asset prices. A DB plan participant’s fore-casted value of plan participation is a function oflength of service, salary, and other factors, suchas vesting provisions, but the driving factor is thelevel of wages. In contrast, given the contributionlevel, the value of a DC plan participation is en-tirely a function of the performance of assets heldby the participant.

Modern finance, as represented by the CapitalAsset Pricing Model and Arbitrage Pricing The-ory (Bodie, Kane, and Marcus 1999), argues thatreturns to financial assets are driven by theirsystematic risks, that is, nondiversifiable, societalrisks assumed by having a position in a givencapital asset. Participation in a DB plan, from theperspective of the plan participant, is a securitythat is purchased with wage concessions duringthe working career and whose payoff is thestream of pension benefits. It could be argued thatthe risks of capital markets are not present in thatsecurity. Indeed, this security can be consideredfree from asset default and equity market decline,as well as interest rate risk.

But it is not free of the risk of fluctuations ofone macroeconomic factor, nondiversifiable toplan participants: the overall growth rate of wagesin the national economy. It should be stressedthat the rate of return to plan participants shouldnot be confused with the internal rate of returnon plan contributions. The plan sponsor assumesa long position in capital assets of the plan andpays for them with plan contributions.

Plan participants pay for their pension benefitswith wage concessions throughout their careers,and their long, invested position is not in the planassets, but in a security whose only uncertainty istied to change in the overall level of wages in thenational economy throughout their working ca-reer. Workers receive individual wage increasesbecause of their individual merit and careerprogress, but these are not systematic, and notfully certain, and from the perspective of modernfinance cannot be priced into the expected re-turns to plan participants. Only the macroeco-nomic factor of national wage growth is a sourceof systematic risk in DB pension plans, and assuch it should determine the rates of return to

plan participants. In this paper private DB plansare studied from that perspective.

Gwartney and Stroup (1995) provide dataabout average productivity and real wage in-creases in the United States in a subset of thepost–World War II period (see Table 2). A generaldownward trend is seen in those quantities. Re-call that the classical economic model of laborand capital as factors of production takes theproduction function as

Y 5 F~K, L, t!, (1)

where Y is the production, K is capital input, L islabor input, and t is time. The function F is typ-ically assumed to be homogeneous of degree one.It is also assumed that partial derivatives ]F/]Kand ]F/]L are positive, while the second partials]2F/]K2 and ]2F/]L2 are negative (the law of di-minishing returns). Surrey (1976) provides aclassical account of the problem of the distribu-tion of the national income between capital andlabor. Note that ]F/]K is called the marginal pro-ductivity of capital, and Y/K is the average pro-ductivity (or average product) of capital; similarly]F/]L is the marginal productivity of labor, andY/L is the average productivity (or average prod-uct) of labor.

The expressions

]F/]K

Y/Kand

]F/]L

Y/L(2)

are called relative shares of capital and labor,respectively, and represent portions of new out-put that are due to the productivity increase ofcapital and labor, respectively. The most standardmodel of the production function is the Cobb-Douglas formula,

Table 2Productivity and Wage Growth

in the United States

Period

Average AnnualProductivity

Growth

Average AnnualReal Wage

Growth

1948–55 3.70% 3.40%1956–68 3.00 3.101969–80 1.20 1.001981–92 1.30 0.60

Source: Gwartney and Stroup (1995).


Y 5 AKaL12a, (3)

where A and a are constant parameters. In thiscase, we have

]F/]K

Y/K5 a and

]F/]L

Y/L5 1 2 a. (4)

If a given level of output is sought for the mini-mum total cost of capital and labor, the well-known equilibrium condition is

Marginal Productivity of LaborPrice of Labor

5Marginal Productivity of Capital

Price of Capital. (5)

In a perfectly competitive labor market, theprice of labor is equal to the marginal productiv-ity of labor (Gwartney and Stroup 1995). It can befound from historical data that these two quanti-ties do follow each other reasonably closely. (Forfascinating evidence of how closely this theoreti-cal result follows the historical reality of wages inthe United States, see Cooper and Borden 1997.)Also observe that productivity increases haveslowed since the 1970s, resulting in a smaller rateof increase (or even in decreases) of real wages.Under the Cobb-Douglas function model of pro-duction, relative shares of labor and capital re-main constant over time. The debate about

whether these do indeed remain constant is oneof the central issues in classical economics. Thelast two decades in the United States have broughtlower rates of increase in real wages, while the realrates of return on capital have increased. Theshares of the two factors of production may appearto have changed their relationship.

Figure 2 shows average real weekly wages(1996 dollars) in the United States in January ofeach year between 1964 and 1996, based on dataprovided by the U.S. Department of Labor. Thedata show a slightly declining trend, definitely notcorresponding to the overall growth of the grossdomestic product and the national economy inthat period.

Figure 3 addresses another important issue. Itshows the U.S. wages and salaries index, as pro-vided by the Department of Labor, seasonallyadjusted and in nominal values compared withnominal values of an amount equal to the initialvalue of the index invested in U.S. Treasury bills.The rate of return on Treasury bills has signifi-cantly exceeded the rate of growth of wages.

This means that over the 16 years presented inthe graph investing in the wages and salaries in-dex would have resulted, on a national average, inabout half of the amount accumulated when com-pared to investing in Treasury bills. It is oftenargued that DC plan participants take too littlerisk with their assets, often placing them inmoney market instruments or Treasury bill

Figure 2U.S. Real Weekly Wages (January)

Source: U.S. Department of Labor.

Figure 3Investing in Treasury Bills Versus Investing

in Wage Growth



equivalents. Yet, if the data are an indication ofthe existing macroeconomic patterns, partici-pants may not be reducing their risk by purchas-ing those money market instruments, but actu-ally improving their returns if they view their DCinvestments in comparison with participation inDB plans.

Of course, the overall growth rate of wages isnot the only determinant of an individual DBpension level. There are other factors, such asindividual performance, vesting provisions, andservice with an employer. But if these latter fac-tors are not systematic, we should expect a typi-cal predictable pattern of merit and seniority in-creases to be priced into the bargaining process ofwage concessions paid in return for pension ben-efits. As was stressed previously, if DB plan par-ticipation as a security is compared to a DC planas a security, DC is merely a perfect conduit ofunderlying asset performance, while DB partici-pation is a derivative security creating wage-de-pendent cash flows out of the underlying portfolioof capital assets. If this DB derivative security ispriced according to modern finance theory, wewould expect it to be priced based on the system-atic factor of national wage growth.

In his analysis of social security systems (suchas the U.S. OASDI, which can be considered auniversal DB plan) Samuelson (1958; see also1975) developed a model in which a steady-statesocial security system delivers to its participantsin the accumulation phase a nominal rate of re-turn given by the compounding of population andlabor productivity growth (assuming that laborproductivity is fully reflected in taxable payrolland assuming full compliance). The same reason-ing naturally applies to DB plans, if taken in ag-gregate. This is very much in line with financialpricing of DB plan participation. However, in thecase of DB plans, only labor productivity is trans-lated into the security performance of DB partic-ipation (social security systems have the advan-tage of compulsory participation with respect tothe entire population, thus gaining the possiblebenefit of demographic growth, tax increases, andincreases in population coverage). The data inTable 2 show that labor productivity grew at realrates over 3% during the period 1948–68, makingDB participation a relatively attractive security,while the last three decades have witnessed a

dramatic change in the rate of growth of laborproductivity, dropping below 1% real.

As Robert C. Merton says in his discussion ofthe work of Diamond and Mirrlees (1985), “In aworld of full information and perfect markets,where all assets (including human capital) arefreely tradable, private pensions provide nothingmore than another way for individuals to save.With a full complement of risk-sharing securitiesavailable, the worker can fully offset or modifyany particular form of payout prescribed by apension plan” (p. 357). This would suggest that insuch a world the workers would hedge their po-sitions in DB plans and create an optimal portfolioof various capital assets, possibly including DBplan participation. Why then did DB plans de-velop in the first place, and why are we nowwitnessing such a large shift to DC plans?

Some insights into the economic meaning ofDB plans are provided by the analysis of marketimperfections embedded in the labor market. Thekey issue is that human capital is not tradable.Young workers have large amounts of human cap-ital but cannot trade it for any other instruments.This forces such young workers to save more thanan optimal amount (as any savings on their partmay be excessive, given a large human capitalposition). Ideally they should seek ways to dis-save and, by doing so, pursue diversification ofhuman capital and the resulting reduction of non-systematic risk (i.e., nonmarket risk of their ownhuman capital). One way to dissave would be toaccept lower wages, or lower pension contribu-tions, in return for the purchase of items of valueto young workers.

Diamond and Mirrlees (1985) show that opti-mal labor contracts with nontradable human cap-ital contain a portfolio of options, such as theworker’s option to seek other employment. Thisperspective can be simplified by stating thatyoung workers purchase their mobility by givingup some wages. This can be achieved more effi-ciently by allocating very small amounts to theirpension contributions, instead of directly affect-ing their wages. One good way to achieve thatpurpose is through a promise of a DB plan, butwith slow vesting.

Bulow and Scholes (1983) provide yet anotherperspective on the DB plans. They point out thatthe vesting option can be viewed as being tradedbetween young employees, who purchase it, and


elderly employees, who hold it and gradually re-sell it. This results in some compensation ofyoung employees being allocated to vesting op-tion purchase.

Thus, under DB plans, participants can beviewed as holding two types of securities: (1) DBplan participation, whose performance is linkedto labor productivity, or wage growth at the mac-roeconomic level, and (2) various options, includ-ing the option to vest, retire early, and so on. Incontrast, DC plan participants hold securitiesthat are really the underlying capital assets. It iswell known that large amounts of DC assets in theUnited States are held in low-risk securities, suchas money market instruments. Yet in the last twodecades even those low-risk securities outper-formed the DB plan participation securities, thusgiving even very risk-averse DC plan participantshigher returns for a comparable level of risk. Themain hypothesis here is that the historical shift toDB plans represents a lifetime investment portfo-lio readjustment on the part of participants. Itleaves workers better off in terms of returns, andit appears not to be to their detriment in terms ofrisk. Indeed, given greater uncertainty of employ-ment (expressed, e.g., in downsizing by manylarge American corporations), a hypothesis canbe ventured that some improvement in their riskprofile was achieved. However, the claims ofgreater uncertainty of employment can be uncer-tain itself. On one hand, perception of uncer-tainty, due to downsizing and restructuring,seems to exist among the general public. On theother hand, unemployment had been trendingdown and is now at historical lows, and empiricalstudies do not confirm the perception of employ-ment uncertainty.

As is the case with all macroeconomic pro-nouncements, our explanation of the DC plans’expansion at the expense of DB plans is not easyto prove empirically. Controlled experiments arenot possible in the national economy. There aremany factors that influence the pension behaviorof both employees and employers. One could ar-gue that all pension-related decisions are madesolely by employers. This paper disagrees withthat perspective, viewing it as too extreme. Aspointed out in several analyses in Wise (1985),especially in Wise’s own overview, and in Lazear(1985), DB plan participation is strongly associ-ated with union membership and with work in

large corporations. Union membership has beenin decline in the last two decades, and this to agreat degree can be attributed to employees’choices. Also, workers downsized from large cor-porations do not necessarily pursue employmentin other large corporations and may do so bychoice. They often opt for self-employment anduse DC plans. The 1997 introduction of SIMPLEDC plans for small businesses reflects the growingneeds of that portion of the political constituency.

Therefore, it is proposed that, in the face ofhigher returns of capital assets in DC plans, withthe same or lower level of risk as the DB planparticipation derivative security, a shift to DCplans is a movement to a superior security, whosesuperiority is a function of the decline of theimportance of wages in the national income. Thispaper will now attempt to provide an empiricaljustification for this hypothesis.

4. ASSESSING THE HYPOTHESIS

Note that, given the macroeconomic nature of thedata, the impossibility of performing controlledexperiments, and the fact that we are workingwith nonstationary time series, the power of sta-tistics is limited when dealing with questionssuch as this one. Nevertheless, the statistical re-lationship of returns to labor versus participationin DB plans, if one exists, may shed some newlight on the changes in the relative position of DBand DC plans in the recent history of the UnitedStates. To examine such a relationship, wagesand DB plans must be considered in an aggregatedperspective. If individual wages are not growing,but the share of national income to wages is un-changed, then maybe there are too many workersfor significant increases in salaries to occur. Butsuch a situation typically would be temporary. Iam proposing that a meaningful measure of thesignificance of wage growth is its growth in rela-tion to national income, that is, the share of na-tional income flowing to wages. Thus I have de-signed a test of my hypothesis of the causalrelationship between the falling relative impor-tance of wages in national income and the fallingrelative importance of DB plans in pensions. Thisis done by comparing several data series.

First, let us compare (1) the portion of all re-tirement plan assets (or new contributions, ben-efits, and so on) in DB plans, and (2) the portion


of total GDP attributable to labor. The first dataseries can be derived from data published by thePension and Welfare Benefits Administration ofthe U.S. Department of Labor (2000), and thesecond data series can be derived from NationalIncome and Product Accounts as published bythe Bureau of Economic Analysis of the U.S. De-partment of Commerce (2000).

The National Income and Product Accountsexpress the gross national product of the UnitedStates from the income side by providing infor-mation about total employee compensation, in-cluding wages and salaries and nonwage compen-sation (such as benefits and social insurancecontributions), proprietors’ income, and incomefrom capital such as rents, dividends, interest,and corporate income. This national accountingprocedure allows for an integrated view of incomesources for Americans and, from the perspectiveof this paper, is a valuable source of attribution ofincome to the factors of labor and capital. Thehypothesis will be analyzed by regressing the por-tion of pension assets allocated to DB plans on theportion of the national income flowing to labor,and, in the second analysis, by analyzing the por-tion of new pension contributions flowing to DBplans on the portion of the national income flow-ing to labor. All of the data are annual, as this isthe way pension data are provided by the Depart-ment of Labor.

This approach may not fully resolve the prob-lem of the relationship of the two variables (DBpension contributions, or assets, and wages). Asalready pointed out, a macroeconomic approachto pensions brings with it many other questionsconcerning factors contributing to the changes ofthe variables analyzed. Furthermore, the time se-ries analyzed are nonstationary and, thus, mayrequire additional analysis, beyond the scope ofthis paper, concerning their time-dependent be-havior. However, I believe that this contributionprovides a new perspective that has been largelyignored in actuarial literature.

Figure 4 shows the portion of national incomeattributed to wages and salaries versus the por-tion of pension assets in DB plans from 1975 to1995. It can be argued that wages and salaries donot constitute all of employees’ compensation.Because of this concern, the shares of nationalincome to total compensation of employees, arealso included as well as the share to total com-

pensation of employees and to sole proprietors(this last quantity representing the widest possi-ble measure of national income attributable tolabor). However, from the perspective of this pa-per it is more appropriate to analyze only wagesand salaries as labor’s income, as other compen-sation typically does not affect current pensioncontributions and employee incentives (especi-ally for DB plans). Note that the most steeplydeclining curve represents the portion of pensionassets in DB plans and that we do not have officialDepartment of Labor data on this data series be-yond 1996, while the wages/salaries portion of thenational income graph extends through 1998,showing a possibly historic upturn starting in1997 (not yet compared to the DB portion ofpension plans in this analysis).

While the top line in Figure 4, which shows theportion of national income flowing to total com-pensation and proprietors’ income, does not showany significant declining trend, there are someapparent dependencies. For further investigation,only the bottom two lines are repeated in Figure5: national income share paid in wages and shareof pension assets in DB plans. As indicated ear-lier, wages are the basis for DB benefit level, andthey determine long-term returns for DB planparticipants.

We can see the striking difference between theattribution of the share of national income flow-

Figure 4Shares of National Income to Wages, Total

Compensation, and Total Compensation PlusProprietors’ Income Versus Share of Pension

Assets to DB Plans



ing to labor by wages and salaries only and theshare provided by wages and salaries plus benefitsand the share including proprietors’ income inFigure 4. While wages and salaries have beenmostly in a downward trend established in the1970s, the other two measures do not appear tohave such a clear trend. As already noted, thesetwo wider measures of labor compensation aretypically not included in the wage basis for thecalculation of benefits in DB plans. The relativeattractiveness of DB plan participation as a secu-rity is determined by the growth of what can betermed pensionable earnings, which is repre-sented by wages and salaries.

Figure 5 shows that the share of national in-come flowing to wages and salaries appears to becorrelated with the share of pension assets in DBplans. In Figure 6 we see that there is a verystrong relationship between the two variables.The linear model is highly statistically significant(p-value of 0.01%), with R2 of over 60% and anadjusted R2 of over 64%. Both parameters of linearregression are equally highly statistically signifi-cant. The share of national income flowing towages appears to provide some degree of explan-atory power for the share of pension assets in DBplans.

We may ask, of course, if we should look at newmoney flowing into DB plans instead of total as-sets. The response on the part of employees to theeconomic incentive is probably better repre-

sented by new contributions, but assets are animportant part of the picture too, as they repre-sent the wealth of pension plan participants.

Figure 7 shows historical patterns of the shareof national income flowing to wages and salariesand new contributions to DB plans as a portion ofall pension plan contributions. This is followed bya linear regression in Figure 8. Results indicate aneven stronger statistical significance of the rela-tionship, with similar values of R2 and adjusted R2

as in the case of share of pension assets in DBplans.

Figure 5Share of National Income to Wages Versus

Share of Pension Assets to DB Plans


Figure 6Analysis of Share of Pension Assets in DB Plans

Versus Share of National Income Flowingto Wages, 1948–1968

Figure 7Share of National Income to Wages VersusShare of Pension Contributions to DB Plans

Source: Department of Labor.


A hypothesis may be ventured that employeesare not as quick to respond to economic incen-tives by changing their pension contributions,their employment contract, or their job as simul-taneous analysis of the variables used would im-ply. This can be addressed by considering therelationship of the share of DB assets and contri-butions as a function of the previous year’s shareof national income flowing to wages, or the yearbefore last, or even further back. An alternativeapproach would be to study the first differenceseries. However, such an analysis in the casestudied here does not indicate any change in thestrength or the nature of the relationships.

To illustrate that, Figure 9 shows the results ofan analysis of the share of the national incomeflowing to labor versus DB assets and DB contri-butions lagged by one year. Statistical signifi-cance does indeed improve, but only slightly. Re-sults for other variables and other time lags, aswell as first differences, are similar. Since theprocesses involved here are long-term macroeco-nomic phenomena, there may be lags, but theyshould not affect greatly the underlying economictrends, and in all cases, with or without lags, wecan see the statistical significance of the models.

Note also the following:

● The share of DB plan assets in all plans has a93.41% correlation to the share of new pension

contributions going to DB plans. Even thoughthere was a shift in the relationship between thetwo plan types, contributions previously goingto DB plans were not necessarily redirected toDC plans, but rather newly started plans wereapparently mostly of the DC type.

● The share of participant numbers in DB plansalso had a very high correlation with the shareof total plan assets in DB plans, 90.36%.

● The share of benefits paid by DB plans had aneven higher correlation with the total plan as-sets share than with the share of participants, at90.61%.

The implication of these points is that the shiftaway from DB plans appears to have been a com-prehensive process, and as this analysis points out,it is strongly correlated with the decline of the shareof national income flowing to wages and salaries.

Various nonlinear regressions of the same vari-ables as the ones studied above have been per-formed, but nonlinear regression results turned outto be dramatically less statistically significant, withthe exception of odd-power polynomial models,which give close approximations of the linearmodel.

At this point it seems appropriate to ask one morequestion: How did the shift out of DB plans correlatewith the share of national income flowing to thewidely defined compensation of employees? Thetwo time series are shown in Figure 10. Statistical

Figure 9Analysis of Share of DB Assets in Pension PlanAssets Versus Previous Year’s Share of National

Income Flowing to Wages, 1975–1995

Figure 8Analysis of Share of National Income to Wages

Versus Share of Pension Plan Contributionsto DB Plans


analysis of the DB assets and the share to employ-ees’ compensation (not including proprietors’ in-come) appears in Figure 11. We see that the twotime series fail to have any meaningful relationshipto each other.

It has been a long-standing contention amongsome economists (Gwartney and Stroup 1995;Surrey 1976) that the share of national incomeflowing to labor remains more or less constantover time (as it is in the Cobb-Douglas modeldiscussed previously). We have seen from previ-ously presented data that this has not been true ofthe share going to wages and salaries. However, asshown in Figure 10, the share to total compensa-tion (which includes not only wages/salaries butalso the costs of benefits and social insurancecontributions paid by employers) has remainedrelatively stable in the tumultuous years of thelast quarter century, with a mean of 72.31% andthe 95% confidence interval for the mean being(71.96%, 72.66%).

One must conclude, of course, that the last 25years have brought about a change in the rela-tionship between total compensation includingproprietors’ income and wages. Figure 12 showsthe trends in the share of national income thatbelongs to the total compensation of labor but isnot paid wages (i.e., cost of benefits and socialinsurance contributions). This portion of nationalincome has been dramatically increasing, with arather recent downturn.

5. CONCLUSIONS

This paper has shown that a high degree of cor-relation exists between the shift away from DBplans and the nearly equally pronounced, but notmentioned as often, shift away from compensat-ing labor in the form of wages. As the emerging

Figure 11Analysis of Share of Pension Assets in DB Plans

Versus Share of National Income to TotalCompensation of Labor

Figure 12Share of National Income Belonging to TotalCompensation of Labor but Not Derived from

Wages (Benefits and Social InsuranceContributions)

Figure 10Share of National Income to Total Labor

Compensation and Share of PensionAssets in DB Plans

Source: Department of Labor.


forms of compensation do not lend themselveseasily to be the basis for deriving a benefit in a DBplan, it would seem natural that different forms ofpensions serve the workers who receive theircompensation in a different form. It should bestressed, however, that the best we can do at thispoint is to derive correlations and hypothesizecausations. When two random variables, say, Xand Y, are correlated, this can generally indicateone of the following three alternatives:

1. X may cause Y2. Y may cause X3. Both X and Y may be caused by some third

variable, or a combination of variables, thatmay or may not be included in the analysis.

Furthermore, we are not merely studying ran-dom variables here; these are time series. Asthese series are nonstationary, statistical rela-tionships developed here are even more suspect.

Nevertheless, this paper does point to the im-portance of wages as an index for DB plans and tothe weakness of that wage index within nationalincome coinciding with the weakness of DB plans.Actuarial scholars and practitioners may benefitfrom this observation. We can hope that this cor-relation may mean causation. It may also meanthat both wages and DB plans have been affectedby changes in the national economy, the evolvingrole of government, or changes in risk perceptionand distribution.

We could even view this analysis as a vindica-tion of the New Economy Theory. The DB planparticipation derivative security created by DBplans is indexed to wages and salaries, which havedeclined in their relative importance in the na-tional income in this new economy and havegreatly underperformed securities available in thecapital markets. This has created an incentive topurchase other securities, and economic decisionmakers appear to have responded to that incen-tive. Also, workers seem to be paid differently inthe new economy, not necessarily in directwages, but in benefits, including ever more ex-pensive health insurance and forms of participa-tion in capital assets such as DC plans, stockoptions, and, of course, income derived from self-employment in sole proprietorships. This is, infact, another way to express the New EconomyTheory contention.

Does this mean that the Excessive RegulationTheory is entirely false? To the degree it focuseson some form of bad intentions or conspiracy, itis. However, to the degree that government activ-ities create incentives for economic decisionmakers, regulators of pension plans who careabout the future of DB plans would be well ad-vised to take this analysis into consideration. Theimplication here seems to be that one possiblekey to the future of DB plans is to make DB planparticipation security indexation to wages andsalaries attractive again, or to change the methodof calculation of benefit to allow for the inclusionof capital markets’ performance. The case for theRisk-Averse Employers Theory is quite weak,given that even risk-free marketable securities(Treasury bills) have provided higher growth ofcapital than the wage index.

In any case, it must be admitted that there maynot be a complete solution to the puzzle of the DBplans’ decline here, merely a different perspectiveon this important issue. DB pension plans havehistorically given a valuable safety net to Ameri-can workers. It would be undesirable from thepoint of view of public policy to allow them todisappear without providing workers with somesubstitute to meet their needs. DC pensions, asattractive as they appear now, have greater inher-ent uncertainty about the relationship of the pen-sion and the final salary. Receiving insuranceagainst that uncertainty is a benefit that one daymay be desired again by many. This is especiallytimely in view of the changes in the Americaneconomy in the second half of the 1990s, whichseem to indicate a reversal of the historical trendof a decline in the significance of wages in na-tional income. This issue will undoubtedly war-rant further investigation.

ACKNOWLEDGMENTS

This work has been prepared as a research papersubmitted for credit toward Fellowship in the So-ciety of Actuaries. The author expresses his grat-itude to the Research Paper Committee as well asto James Hickman, Ph.D., F.S.A., Shane Chalke,F.S.A., and the anonymous referees for their ad-vice in the preparation of the work. All statisticalanalysis of this work was performed using JMP, astatistical software package for the Macintosh op-erating system.


ReferencesBHARMAL, SHIRAZ Y. M. 1988. “Debate No. 7: Defined Benefit

Pension Plans Are Not Well Suited to Tomorrow’s WorkForce and Should Be Replaced by Defined ContributionPlans.” Proceedings of the Canadian Institute of Actuaries19(2):117–33. Proceedings from the March 7–8, 1988, Spe-cial Meeting of the Canadian Institute of Actuaries.

BODIE, ZVI, ALEX KANE, AND ALAN J. MARCUS. 1999. Investments. 4thed. New York: Irwin McGraw-Hill.

BULOW, JEREMY I., AND MYRON S. SCHOLES. 1983. “Who Owns theAssets in a Defined-Benefit Pension Plan?” in FinancialAspects of the United States Pension System, ed. Zvi Bodieand John B. Shoven, pp. 17–36. Chicago: University ofChicago Press.

COOPER, JAMES C., AND KARL BORDEN. 1997. “The Interpretation ofWages and Prices in Public Historical Displays.” PublicHistorian 19(2): 9–36.

DIAMOND, PETER A., AND JAMES A. MIRRLEES. 1985. “Insurance As-pects of Pensions.” Comment by Robert C. Merton. InPension, Labor, and Individual Choice, ed. David A. Wise,pp. 317–56. Chicago: University of Chicago Press.

GWARTNEY, JAMES D., AND RICHARD L. STROUP. 1995. Economics:Private and Public Choice. 7th ed. Orlando, FL: HarcourtBrace & Co.

LAZEAR, EDWARD P. 1985. “Incentive Effects of Pensions.” In Pen-sion, Labor, and Individual Choice, ed. David A. Wise, pp.253–82. Chicago: University of Chicago Press.

MARCUS, ALAN J. 1987. “Corporate Pension Policy and the Valueof PBGC Insurance.” In Issues in Pension Economics, ed.Zvi Bodie, John B. Shoven, and David A. Wise, pp. 49–81.Chicago: University of Chicago Press.

MERTON, ROBERT C. 1983. “On the Role of Social Security as aMeans for Efficient Risk Sharing in an Economy Where

Human Capital Is Not Tradable.” In Financial Aspects ofthe United States Pension System, ed. Zvi Bodie and JohnB. Shoven, pp. 325–58. Chicago: University of ChicagoPress.

POTERBA, JAMES M., SEAN VENTI, AND DAVID A. WISE. 1994. “401(k)Plans and Tax-Deferred Saving.” In Studies in the Econom-ics of Aging, ed. David A. Wise, pp. 105–42. Chicago:University of Chicago Press.

SAMUELSON, PAUL A. 1958. “An Exact Consumption-Loan Model ofInterest with or without the Social Contrivance of Money.”Journal of Political Economy 46 (December): 467–82.

———. 1975. “Optimum Social Security in a Life-Cycle GrowthModel.” International Economic Review 16 (October):531–38. Reprinted in Collected Scientific Papers of Paul A.Samuelson, vol. 4. Cambridge, MA: MIT Press.

SURREY, M. J. C., ED. 1976. Macroeconomic Themes. Oxford:Oxford University Press.

U.S. DEPARTMENT OF COMMERCE, BUREAU OF ECONOMIC ANALYSIS. 2000.National Accounts Data. Available online at http://www.bea.doc.gov/bea/dn1.htm.

U.S. DEPARTMENT OF LABOR, PENSION AND WELFARE BENEFITS ADMINIS-TRATION. 2000. Private Pension Plans Bulletin, Abstracts of1996 Form 5500 Annual Reports. Available online athttp://www.dol.gov/dol/pwba/public/programs/opr/bullet1996/cover.htm.

WISE, DAVID A., ED. 1985. Pension, Labor, and Individual Choice.Chicago: University of Chicago Press.

Discussions on this paper can be submitted until Jan.1, 2002. The author reserves the right to reply to anydiscussion. Please see the Submission Guidelines forAuthors on the inside back cover for instructions onthe submission of discussions.


naaj vol 5 no 3 july 2001, macroeconomic aspects of private

Documents