The balance-of-payments constraint on economic growth in a long-term perspective: Spain, 1850-2000*

Oscar Bajo-Rubio(Universidad de Castilla-La Mancha)

This version, September 2011

AbstractThe balance of payments can act as a constraint on the rate of growth of output, since it puts a limit on the growth in the level of demand to which supply can adapt. In this paper, we examine this issue for the case of Spain, using time series data extending over the period 1850-2000. Overall, the external deficit does not seem to have worked as a constraint on the growth of the Spanish economy over the long run, unless some shorter and specific subperiods, such as 1940-1959 and 1959-1974. The Spanish economy seems to have used external deficits to smooth her level of aggregate consumption, which would be supported by the finding of sustainability of current account deficits along the period of analysis.

Key words: Economic growth, External deficit, Spanish economy

JEL Classification: F41, F43, N10

________________________* The author wish to thank an anonymous referee for helpful comments to a previous version.

Financial support from the Spanish Ministry of Science and Innovation (Project ECO2008-05072-C02-01), and from the Department of Education and Science of the regional government of Castilla-La Mancha (Project PEII09-0072-7392), is also gratefully acknowledged.

1. Introduction

The Spanish economy has experienced a steady process of growth since the beginning

of industrialization at the start of the 19th century. However, though following a rather

similar evolution to that of the rest of Western Europe, she underwent a relative

retardation as compared to those countries. In particular, the real GDP of Spain

multiplied by 40 over the period 1850-2000, with an annual average rate of growth of

2.4 percent; in per capita terms, GDP multiplied by 15 over the same period, and the

average rate of growth was 1.7 percent per year. But, despite the intense catching-up

that took place in the last fifty years, per capita GDP of Spain at the end of the 20th

century was still around three quarters of Western Europe’s, roughly the same as one

hundred years before (Prados de la Escosura, 2007).

On the other hand, the foreign sector has played a critical role in the

development of the Spanish economy, in a country in which the endowments of natural

resources were rather poor, and traditionally characterized by a relative backwardness in

relation to their neighbouring countries. In addition, the loss of most of the colonial

empire in the first decades of the 19th century reinforced that crucial role of the foreign

sector. The disappearance of the monopoly of trade with the American colonies meant

that exports to the colonies and re-exporting colonial products to Europe were both

drastically reduced, which translated into a reorientation of Spanish exports to the

European markets and a large trade deficit. The assessment of the role played by the

foreign sector in the economic progress of Spain has been controversial, though; see,

e.g., Nadal (1975) or Prados de la Escosura (1988). An account of the main

developments of the Spanish foreign sector in the last two centuries is provided in

Tortella (2000).

The important modernizing role played by the foreign sector on the evolution of

the whole economy, though limited by its small relative size, has been emphasized in

Prados de la Escosura (1988). In fact, the Spanish economy experienced a higher

growth over those periods characterized by a greater external openness (as in, e.g., the

1960s, or the years after 1986); and, likewise, she fell behind coinciding with periods of

a greater isolation against the rest of the world (such as the years 1890-1913, or 1930-

1950) (Prados de la Escosura, 2007). On the whole, the lower weight of foreign trade on

GDP as compared to the “European norm” (Molinas and Prados de la Escosura, 1989)

1

might explain the relative backwardness of the Spanish economy until the last third of

the 20th century.

However, some authors have also argued that the trade deficit had been a

chronic problem of the Spanish economy, given the traditional weakness of exports, and

the need for essential imports such as energy, raw materials, intermediate products, and

equipment goods. This, in turn, would have been translated into a subordination of

growth to the evolution of the trade balance; see, e.g., Segura and García-Viñuela

(1978).

The recent availability of long time series for both GDP (Prados de la Escosura,

2003) and the foreign sector (Tena, 2005, 2007; Prados de la Escosura, 2010) invites to

a re-examination of the relationship between growth and foreign trade for the case of the

Spanish economy. In particular, the aim of this paper will be to investigate whether or

not the balance of payments would have meant an impediment to further GDP growth

over the period 1850-2000. More specifically, we will analyze the two subperiods 1850-

1913 and 1940-2000, on the grounds of availability of data on the current account

balance (see below); leaving aside the interwar years, characterized by the disruption of

the world economic order related to the first globalization. The paper is organized as

follows: section 2 offers a review of the literature on the relationship between external

deficit and economic growth, together with the available evidence for the Spanish case;

in section 3 we investigate whether the external deficit was a constraint to economic

growth, which is complemented with an analysis of the sustainability of current account

deficits in section 4; finally, section 5 concludes.

2

2. Review of the literature

The relationship between international trade (or, more generally, external openness) and

economic growth is a widely discussed topic in economics, but still a controversial one.

Indeed, last years have contemplated a renewed interested on the subject, following the

appearance of the literature on endogenous growth. These theories allow for a greater

role of the countries’ external openness in the process of generation and transmission of

technology, as compared to the traditional, Solow-type, growth models. Specifically, by

developing ideas previously put forward by Schumpeter (1934), the new approaches

emphasize that technological innovation appears in response to economic incentives;

where the institutional, legal, and economic environments, including the extent of

external openness and economic integration, also affect the pace and direction of

technological change. A survey of this literature is provided in Grossman and Helpman

(1994) or, more extensively, in Aghion and Howitt (1998).

On the other hand, there is also a large body of empirical literature based on the

idea that international trade and, in general, a more open commercial policy, means a

major factor in order to explain economic growth. This line of research, indeed, has

been given a renewed impulse with the development of endogenous growth theories, on

giving it some more solid theoretical foundations. In particular, it has been emphasized

that a more liberalized trade stance allows countries to enjoy a higher amount of

intermediate inputs at a lower cost, and encourages technological progress; and all this

would result in higher rates of growth. This literature, which is mostly concerned about

the case of developing countries, is surveyed in Edwards (1992, 1993). The evidence in

favour of this hypothesis, though, is far from being unambiguous, as shown in an

influential paper by Rodríguez and Rodrik (2001). As these authors point out, the

relationship between external openness and economic growth would be rather a

contingent one, dependent on a host of particular characteristics, both country-specific

and external; and, among them, the importance of institutions would be crucial.

Our main concern in this paper, however, will be the analysis of the relationship

between external deficit and economic growth. Is the external deficit really an

impediment to economic growth? Or, maybe, is the external deficit an inevitable

consequence of economic growth for a relatively backward country; and even

3

something necessary in order to catch-up and converge towards the more advanced

countries?

A well-known approach to deal with this issue is due to Thirlwall (1979).

According to this author, the balance of payments can act as a constraint on the rate of

growth of output, since it puts a limit on the growth in the level of demand to which

supply can adapt. An increase in domestic output, by increasing imports, can lead to a

deficit in the balance of payments, which may require either a fall in demand or a real

exchange rate depreciation (i.e., a worsening of the terms of trade) in order to ensure the

sustainability of the external deficit. Accordingly, an unsustainable external deficit

sooner or later requires a correction, which puts a brake on further output growth.

From here, the concept of balance of payments-constrained growth rate, i.e., the

rate of growth of exports divided by the income elasticity of the demand for imports,

follows; see Thirlwall (1979) and Thirlwall and Hussain (1982). By comparing such a

growth rate with that prevailing in a particular country, it would be possible to assess

whether the balance of payments works as a constraint on economic growth in the

country analyzed. This concept, on the other hand, is equivalent to a result derived by

Krugman (1989), who found that countries growing faster face a higher income

elasticity for their exports than for their imports. The rationale behind this approach is

that no country can grow faster than its balance of payments-constrained growth rate for

a very long time, since its ratio of external debt to GDP will reach unsustainable levels

leading to a collapse in international confidence and to an external debt and currency

crisis. Only when the actual growth rate is lower than the balance of payments-

constrained one a country would be able to experience a sustained growth. A historical

background to the approach is provided in Thirlwall (2011); and a collection of papers

on the subject is McCombie and Thirlwall (2004).

An alternative view on the significance of the external deficit follows the so-

called intertemporal approach to the current account, which has become in last years the

orthodox account for explaining the countries’ external imbalances. The starting point

of this approach is that, from the perspective of the national accounts, the current

account equals the difference between savings and investment. From here, and because

savings and investment decisions are based on intertemporal factors (such as life-cycle

4

considerations, the expected returns of investment projects, and the like) the current

account is necessarily an intertemporal phenomenon. The intertemporal model of the

current account originates in the work of, among others, Sachs (1981, 1982), Obstfeld

(1982), and Svensson and Razin (1983); some overviews are provided in Obstfeld and

Rogoff (1995) and Razin (1995).

According to this approach, the current account acts as a shock absorber that

smoothes the time profile of consumption in the face of shocks to output, investment, or

government expenditure (Sachs, 1982). In other words, if the agents expect an increase

in their future income, they will anticipate their present expenditure, by reducing

savings (i.e., smoothing consumption over time) and increasing investment (since its

expected profitability rises); and the resulting current account deficit will be financed

with foreign savings. Moreover, if the expectations on higher future growth are

confirmed, the external debt will be paid more easily.

Following this line of analysis, an influential paper by Blanchard and Giavazzi

(2002) argues that deeper economic integration leads to an increased external deficit in

poorer countries. Since the latter should be characterized by higher expected rates of

return and better growth prospects, integration with a group of richer countries would

translate into both higher investment and lower savings (via higher consumption), and

hence a greater current account deficit in these countries. Even more, the prospects of

convergence as regards the richer countries should favour growth expectations in poorer

countries, contributing additionally to greater deficits.

An extreme implication of the intertemporal approach to the current account

claims that, if the government budget is balanced, a high current account deficit should

not be a matter of concern, provided that savings and investment decisions are taken

optimally by private agents. This is the so-called “Lawson doctrine”, from the British

Chancellor of the Exchequer in the 1980s, Nigel Lawson; see Corden (2007) for a

defence of this view.

However, as pointed out by Edwards (2002), large external deficits should still

be a motive of concern from a policy perspective. Even if a current account deficit does

not lead necessarily to a currency crisis, it might do it when the deficit is very large; in

5

other terms, “what matters is not whether there is a large deficit, but whether the

country in question is running an ‘unsustainable’ deficit” (Edwards, 2002, p. 29).

In an important contribution, Milesi-Ferretti and Razin (1996) discuss the usual

notion of sustainability in relation to the country’s intertemporal solvency, that is, when

the present discounted value of future trade surpluses equals current external

indebtedness. Put in other words, current account sustainability would be defined as the

ability of an economy of satisfying its intertemporal constraint in the long run, in

absence of a drastic change either in the behaviour of the private sector or in economic

policy (Taylor, 2002). In general, a current account deficit in excess of 5% of GDP is

regarded as unsustainable, so that above this threshold the adjustment process of the

current account usually begins (Freund, 2005).

Turning to the case of Spain, after the commotions associated with the loss of

the American colonies, the Spanish foreign trade grew during the second half of the

19th century faster than in France or Britain (Tortella, 2000). This situation began to

change in the 1890s, however, when a more protectionist policy stance was

implemented. The degree of openness, which had increased steadily since 1850

(although at still lower levels than the European average), began to decrease after 1895,

which was intensified after the Spanish Civil War, and reached its minimum in the

1940s (Tena, 1995). In fact, the latter period, characterized by the pursuit of “autarky”

by the new government issued from the Civil War, represents, not only the minimum in

the ratio of openness, but also the lowest point in the process of convergence of the

Spanish economy towards Western Europe. Although such a trend to the isolation of the

Spanish economy from world markets began to be reverted since the 1960s, external

openness and economic liberalization proved to be still limited (Prados de la Escosura

and Sanz, 1996). Only after joining the now European Union (EU) in 1986 the ratio of

openness reached comparable figures to those of the rest of Western Europe, and the

Spanish economy can be considered as having definitely adopted an institutional

framework comparable to that of the rest of her new partners.

On the other hand, since the classical work of Sardà (1948), the traditional vision

of the Spanish foreign sector has been that of a chronic deficit in the trade balance,

which was even aggravated in higher growth periods. From here, some authors have

6

argued that such a chronic trade deficit would have hindered the economic growth

prospects of the Spanish economy; in other words, the trade deficit would have worked

as a constraint on economic growth.

The empirical evidence on this issue is not too abundant, though. Herranz-

Loncán and Tirado-Fabregat (1996) estimate export and import functions for the period

1870-1913. Since, when looking at the estimated income elasticities for exports and

imports, the former was found to be smaller than the latter, the authors conclude that the

foreign sector would have meant a constraint on economic growth over the period.

Following a related line of analysis, Serrano-Sanz (1997) computes the balance

of payments-constrained growth rate proposed by Thirlwall (1979) (see the next

section) for the periods 1870-1891, 1892-1935, 1940-1959, and 1960-1985. With the

only exception of the first period, in the rest of cases the actual growth rate was above

its balance of payments-constrained counterpart, concluding again that the foreign

sector would have played a limitative role on growth after 1892. Notice that these

results would imply that the foreign sector began to restrict growth just when more

protectionist policies were implemented; see below.

However, this same approach has been applied by Alonso (1999) to the more

recent period 1960-1994. Since he finds a parallel evolution of both the actual and the

balance of payments-constrained growth rates, the external constraint would have not

been binding during that period.

Finally, in a recent paper, Prados de la Escosura (2010) builds series for the

current account balance over the years 1850-1913. The results from the new series

reveal the existence of two broad periods in the evolution of the current account: one of

persistent deficits between 1850 and 1890; and other where surpluses prevailed,

between 1891 and 1913 (with the exception of the years 1899-1904). Since periods of

higher (lower) growth went together with current account deficits (surpluses), the author

concludes that the external constraint would have not worked during that period.

According to Prados de la Escosura, economic growth at the end of the 19th century

was encouraged by the high levels of foreign net capital inflows, which helped to

finance current account deficits and complemented domestic savings. In turn, reversals

7

of net capital inflows (in the form of “sudden stops”, i.e., significantly and

unexpectedly) after 1891 slowed down growth since investment had to rely just on

domestic savings.

In the next section, we will analyze the role of the external constraint on Spanish

economic growth for the period 1850-2000, by estimating export and import elasticities,

and computing the balance of payments-constrained growth rate. As can be seen from

above, the (scarce) available evidence on the subject does not look too clear-cut. Notice

also that the results of some of the papers quoted before, concluding that the foreign

sector would have restrained economic growth, are not free of problems, which could

justify a further analysis of this issue. In particular, as noticed by Prados de la Escosura

(2010), the estimates of the export and import elasticities in both Herranz-Loncán and

Tirado-Fabregat (1996) and Serrano-Sanz (1997) are rather dubious, since they use data

on exports, imports, and prices that do not correct for the errors present in official trade

figures; see Prados de la Escosura (2010, notes 5 and 6) for details. Also, Serrano-Sanz

(1997) makes use of different estimation methods for the first, and for the rest of

subperiods he considers. Unlike these previous studies, we will draw on a generally

accepted dataset for a more extended time interval, focusing on the periods

characterized by a greater openness of the world economy, examining different

alternative episodes within the whole period, and using a common estimation method

for all subperiods.

8

3. Was the external deficit a constraint to economic growth?

Our starting point will be the standard demand functions for exports and imports as

described by (see Goldstein and Khan, 1985):

X=X (Y ¿ ,Q ) ∂X∂ Y ¿>0 , ∂ X

∂ Q<0

M=M (Y ,Q ) ∂ M∂Y

>0 , ∂ M∂Q

>0

where X and M stand for exports and imports volumes, and Y* and Y for foreign and

domestic real output, respectively; Q is the real exchange rate, measured as the price of

domestic goods relative to foreign goods. To close the model, we add the equation for

the balance of payments:

B=X− MQ

+F

where B and F denote, respectively, the balance of payments and net capital inflows in

real terms, in terms of domestic goods.

In order to keep unchanged the balance of payments, it is required that:

dBdt

=X (ε X ,Y ¿ Y ¿+εX , Q ^Q )− MQ (ε M ,Y Y +εM , Q ^ Q−Q )+F F=0

where ε X , Y ¿ , ε M , Y , ε X ,Q , and ε M ,Q are the income and price elasticities of exports and

imports, respectively; and the symbol ^ over a variable denotes its growth rate.

Assuming that initially B = 0, so that X+F= M

Q , in order to have the balance of

payments in equilibrium we must have:

(1−φ ) ε X ,Y ¿ Y ¿−εM , Y ^ Y +[1+(1−φ ) ε X ,Q−ε M ,Q ] ^Q+φ F=0

where φ= F

M /Q .

Assume for a moment there are no capital inflows, so that φ = 0 and

ε X ,Y ¿ Y ¿−ε M ,Y ^Y + (1+ε X ,Q−ε M ,Q ) ^ Q=0

As can be seen from the above equation, a country growing relatively faster (i.e., when

Y >Y ¿) should have, other things equal, a depreciating real exchange rate (provided

9

that the Marshall-Lerner condition holds, i.e., −ε X , Q+ε M ,Q>1 ). However, as shown by

Krugman (1989), in a context of imperfect competition those countries showing higher

growth rates will increase their shares of world markets, not by reducing the relative

prices of the goods they produce, but by raising instead the number of varieties.

Accordingly, such countries will enjoy more favourable income elasticities (i.e., a

higher ε X , Y ¿ and a lower ε M , Y ) and would be able to experience relatively higher

growth rates without the need of real exchange rate depreciation in the long run.

Hence, if there is no long-run trend in the real exchange rate, Q=0 , we would

have the following condition:

ε X ,Y ¿

ε M ,Y=

YY ¿

i.e., a country growing relatively faster should have a relatively higher income elasticity

for exports than for imports; this is Krugman’s (1989) “45-degree rule”. From here, we

can get a related result, by finding the balance of payments-constrained growth rate, Y B ,

i.e., the maximum growth rate a country can achieve while keeping in equilibrium the

balance of payments, and provided that the real exchange rate remains unchanged:

Y B=ε X , Y ¿

ε M ,YY ¿

This rule, derived by Thirlwall (1979), is the dynamic analogue of the Harrod trade

multiplier (Harrod, 1933), and implies that a country growing above Y B will run an

external deficit, which would harm its future growth prospects; conversely, a country

growing below Y B will run an external surplus.

Finally, if the real exchange rate were not constant over the long run, the balance

of payments-constrained growth rate would be:

Y B=ε X , Y ¿ Y ¿+(1+ε X ,Q−ε M ,Q ) ^ Q

εM , Y

which, in the most general case where capital movements are allowed, becomes

(Thirlwall and Hussain, 1982):

10

Y B=(1−φ ) ε X , Y ¿ Y ¿+ [1+ (1−φ ) ε X ,Q−εM , Q ] ^ Q+φ F

ε M ,Y (1)

In the rest of this section, we will provide estimates of the balance of payments-

constrained growth rate for the Spanish economy over the periods 1850-1913 and 1940-

2000, computed from equation (1).

In order to apply the above theoretical framework, we mostly rely on the foreign

sector data of Tena (2007). The period of analysis is 1850-2000 or, more specifically,

the two subperiods 1850-1913 and 1940-2000. We use data on exports and imports,

both in real terms (Table 3, series 9 and 10); real exchange rate, measured as the

quotient of the price indices of exports and imports (Table 4, series 13); and (minus) the

current account balance, as a proxy of capital inflows (Table 8, series 74). Notice,

however, that the latter variable is available from a regular basis only after 1940; for the

years before, we have relied on Prados de la Escosura (2010), who provides series for

the current account balance from 1850 to 1913 (Appendix 1, column 12). The data on

real GDP have been taken from Prados de la Escosura (2003). Finally, regarding foreign

output, the series on real GDP of the EU built by Carreras and Tafunell (2004, 2005)

has been used, as a proxy for the GDP of Western Europe. Notice that Western Europe

was the main market for Spanish exports over the whole period, with a share on total

exports always above 50 percent (Tena, 2007).

First of all, we have estimated the exports and imports elasticities using the

method of Phillips and Hansen (1990). This method tries to eliminate the potential

biases that could appear when estimating under ordinary least squares, by computing a

class of Wald tests, modified by semiparametric corrections for serial correlation and

endogeneity bias. The values of the estimated elasticities for the periods 1850-1913 and

1940-2000 appear in rows one and four of Table 1, respectively. Notice that the figures

in parentheses are the Phillips and Hansen’s fully-modified Wald test statistics on the

null hypothesis that the estimated coefficients are equal to zero, asymptotically

distributed as a χ2 with one degree of freedom. In addition, rows one and four of Table 2

show the annual average growth rates of the GDP of Spain and the EU, and the balance

of payments-constrained growth rate according to equation (1), for the periods 1850-

1913 and 1940-2000, respectively.

11

[Table 1 here]

[Table 2 here]

As can be seen in Table 1, over the period 1850-1913 the estimated income

elasticities for both exports and imports are rather similar in size. In turn, the real

exchange rate elasticity is only significant for exports (although the coefficient is

estimated with the wrong sign), but not for imports. As a result, the annual average rate

of growth of the Spanish GDP shows in Table 2 a value not too different to the balance

of payments-constrained growth rate for that period.

On the other hand, for the period 1940-2000 the estimated income elasticity for

exports is higher than in the case of imports, whereas the real exchange rate elasticities

are not significant (although the elasticity for imports is significant at the 10% level).

Moreover, the actual and balance of payments-constrained growth rates look rather

similar too, even though somewhat higher for the former. Accordingly, the foreign

sector does not seem to have worked as a constraint on the growth of the Spanish

economy over the long run, along the two subperiods 1850-1913 and 1940-2000.

However, the two periods analyzed are long enough (more than 60 years each) to

hide a potential different behaviour of the series within every period. Hence, in order to

have a more precise picture, we have repeated the above exercise across a set of

different subperiods.

Regarding the period 1850-1913, we have allowed for a possible break point in

1891, which relates to the inward-looking policy stance adopted at the end of the 19th

century. This was characterized by a combination of protectionism of the domestic

industry, the preservation of the domestic market to domestic production (helped by a

strong government interventionism), and economic nationalism (in the sense of

replacing foreign capital for domestic capital as the main driving force of growth). The

changing point of this new policy stance can be dated at the approval of a new and

extremely protectionist tariff in December 1891; see Tena (2006) for a characterization

12

of protectionist policies in Spain at the end of the 19th century and the beginning of the

20th.

The results for the two subperiods 1850-1891 and 1892-1913 appear in rows two

and three of Tables 1 and 2. As shown in Table 1, the estimated income elasticities for

1850-1891 are much higher than for both 1892-1913 and 1850-1913, well above two.

The exchange rate elasticities, in turn, are not significantly different from zero. Again,

the actual growth rates are only slightly above the balance-of-payments constrained

ones, so the foreign sector would not seem to have constrained growth in both

subperiods.

Turning to the period 1940-2000, several alternative break dates have been

allowed for.

A first option would be looking at the evolution of the GDP series, and trying to

detect any possible structural changes in economic growth. We will rely here on the

results of Prados de la Escosura (2003) for the whole period 1850-2000. This author

performs unit root tests including several possible exogenous breaks, and selects those

that maximize the t-statistic of the coefficient on lagged GDP. His results show two

structural changes in the trend of the GDP series dated at 1951 and 1975, together with a

break in the level at 1936. This, on the other hand, leads us to drop the 1940s from the

period 1940-2000, given the poor growth performance of this period in the Spanish

case1 and the deficient degree of internationalization of the world economy, which only

began to resume in the 1950s. In fact, pre-war GDP levels in Spain were not recovered

until 1951; and, in the case of per capita GDP even later, in 1955 (Prados de la

Escosura, 2003). Accordingly, we have calculated the actual and balance of payments-

constrained growth rates for the three subperiods 1951-2000, 1951-1974, and 1975-

2000. The results appear in rows five, six and seven of Tables 1 and 2.

Beginning with 1951-2000, things do not seem to change very much as

compared with 1940-2000 in terms of the estimated elasticities; however, despite being

significantly higher than that of the EU, the average annual rate of GDP growth is now

1 In a recent paper, Prados de la Escosura, Rosés and Sanz-Villarroya (2011) stress the continuity between the policy stance in the 1950s and 1960s, so that the gradual liberalization operated during the 1950s would appear as a precondition for the implementation of the Stabilization Plan in 1959.

13

below the balance of payments-constrained one. When the analysis is repeated for 1951-

1974 and 1975-2000, the estimated income elasticities turn to be much higher,

especially for exports, in the second subperiod (becoming “more favourable”, according

to Krugman’s terminology). Moreover, compared with the balance of payments-

constrained growth rates, the actual rate of growth (always above the EU’s) is now

roughly the same in the first subperiod, and stands slightly below in the second;

suggesting that the foreign sector would have not restrained GDP growth during those

years. These results would emerge following a process of increasing external openness

and structural change, especially intense during the 1960s and reinforced after joining

the EU in 1986. Thus, during the 1950s and 1960s, manufactures began to take the

leading role within the structure of the Spanish foreign trade, previously dominated

mostly by agricultural products so that, at the end of the period, the Spanish pattern of

trade is comparable to that of her EU partners.

An alternative way of characterizing different subperiods would be looking at

the evolution of the foreign trade series. Serrano-Sanz, Sabaté-Sort and Gadea-Rivas

(2008) apply the Bai and Perron (1998) tests for the presence of multiple endogenous

structural changes, to the Spanish foreign sector series over the period 1869-1999,

finding a break at some point in the 1940s. Hence, they identify two stages in the

evolution of the Spanish foreign sector: a first one before 1935, which they call of

“unpredictable trade” (characterized by the predominance of primary products), and a

second one after 1960, which they call of “mature trade” (characterized by the

predominance of manufactures and services); with the 1940s and 1950s being a

transition period. In this way, we have repeated our previous exercise for the two

subperiods 1940-1959 and 1960-2000; and the results are shown in rows eight and nine

of Tables 1 and 2.

As can be seen, for 1940-1959 income elasticities are very small (especially that

of exports), the exchange rate elasticity of imports is wrongly signed, and the actual

growth rate is 1.3 points above the balance of payments-constrained one. In turn, for

1960-2000, income elasticities are quite higher, the exchange rate elasticity of exports is

not significantly different from zero, and the actual growth rate (fairly above that of the

EU) is below its balance of payments-constrained counterpart. Accordingly, for 1940-

1959, i.e., the “autarky” period, the foreign sector might have proved to be a constraint

14

on GDP growth. This result would match the abrupt end of the short-lived growth

period of the 1950s, which led to a high inflation and an increasing external deficit. The

latter, in turn, translated into the almost total depletion of foreign exchange reserves by

mid-1959, which forced General Franco’s regime to a complete reorientation of

economic policies following the advice of the major international economic

organizations (Sardà, 1970). The reconstruction effect, which proved to be very

important for other European countries following World War II (Dumke, 1990), did not

work in the Spanish case. This would reflect the essential role played by institutional

factors in order to sustain growth and catching-up (Crafts, 1992), so that Spain had to

wait until the next decade to enjoy a long-lasting period of growth. In a few words, the

inward-looking policies followed throughout the first two decades of Franco’s

dictatorship, while not preventing growth during the 1950s, proved to be unsustainable

over the long run (Prados de la Escosura, Rosés and Sanz-Villarroya, 2011).

Finally, we have also tried two possible break dates within the period 1960-2000

(the “mature trade” period of Serrano-Sanz et al., 2008), i.e., 1974 and 1985. The results

here should be taken with care, however, given the shortness of the two subperiods

analyzed: just 15 observations for 1960-1974 and 1986-2000; for this reason, we have

not studied the (less interesting) subperiod 1975-1985, involving just 11 observations.

In particular, the years 1960-1974 did contemplate the highest growth rates in modern

Spain. The increased openness of the economy, following the Stabilization Plan of 1959

that ended the autarchic policy stance dominating since the end of the Spanish Civil

War, has been deemed as being at the root of these developments (e.g., Donges, 1971,

1976). The early 1970s meant the end of the expansionary phase experienced by the

Western economies after World War II; so that, after some years of economic stagnation

(that coincided in the Spanish case with the years of transition to a democratic regime),

growth soared again after 1986, the year in which Spain joined the EU. The nature and

effects of trade liberalization policies during those years have been analysed in De la

Dehesa, Ruiz and Torres (1991) for the 1960s and early 1970s; and in Bajo-Rubio and

Torres (1992) for the period following Spain’s entry into the EU. The results for the

subperiods 1960-1974 and 1986-2000 are shown in the last two rows of Tables 1 and 2.

Looking at Table 1, we can see that income elasticities show high values, even

more for exports, and almost twice as great in 1986-2000 as compared to 1960-1974;

15

exchange rate elasticities, however, are not significantly different from zero, except for

imports in 1960-1974. In turn, the actual growth rate is 1.8 points above the balance-of-

payments constrained one in 1960-1974, but below it in 1986-2000.

Therefore, for the period 1960-1974 the foreign sector might have worked as a

constraint on the growth of the Spanish economy. All along this period, the Spanish

economy experienced the strongest growth rates in all her history, well above the EU

average, mostly driven by an impressive increase in total factor productivity (Prados de

la Escosura and Rosés, 2009). Those years were characterized by the recurrent

appearance of situations of “stop-and-go”: when the economy grew too fast the balance

of payments deteriorated, putting pressure on the exchange rate so foreign reserves fell,

and a period of slower growth followed in order to correct the external disequilibrium.

Such a state of affairs would have been also related to the structural transformations of

the 1960s, with an increase in the needs of imported intermediates due to a change both

in the composition of exports and in the entire productive structure (Fanjul and Segura,

1977).

Next, in order to check the robustness of our results to the particular proxy for

foreign output used before (i.e., Carreras and Tafunell’s), we have repeated the analysis

in Table 2 using two alternative measurements of foreign output. Specifically, we have

used (i) the GDP of 12 Western European countries (namely, Austria, Belgium,

Denmark, Finland, France, Germany, Italy, the Netherlands, Norway, Sweden,

Switzerland, and the UK), and (ii) a weighted average of the GDPs of the six biggest

European markets for Spanish exports (namely, Belgium-Netherlands, France, Italy, the

UK, Germany, and Portugal). The GDP data have been taken from Maddison (2010),

and are available on a common basis after 1870; the exports shares on the different

markets come from Tena (2007, Table 6). The six countries chosen represent roughly

more than 70% of Spanish exports during 1870-1913, and somewhat less during 1940-

2000, with a minimum of a bit less than 45% between 1945 and 1960, but reaching

similar levels to those of 1870-1913 once Spain joined the EU in 1986. The new results

are shown in Table 32.

2 In order to save space, we do not show the estimated elasticities from which the balance-of-payments constrained growth rates in Table 3 have been computed. These estimates are available from the author upon request.

16

[Table 3 here]

As can be seen from the table, the new balance-of-payments constrained growth

rates are quite similar to those appearing in Table 2 (recall that, in the first two rows, the

starting year is now 1870 instead of 1850)3. The only noticeable difference is that the

rates are now lower when the 1940s are included in the period of analysis, especially

when the weighted average of the six trading partners’ GDP is used. Accordingly, the

results about the role of the foreign sector as a constraint on growth are virtually

unchanged. Again, the only subperiods in which the actual growth rate is significantly

above the balance-of-payments constrained one are 1940-1959 (1.6 and 2.6 points,

respectively, for each of the two proxies for foreign output) and 1960-1974 (1.8 and 2.2

points).

3 The correlation coefficients with respect to the series of the EU’s GDP of Carreras and Tafunell are 99.9% for the GDP of 12 Western European countries, in both periods; and 98.5% and 99.4% for the weighted average of the six trading partners’ GDP, in the periods 1870-1913 and 1940-2000, respectively.

17

4. Were current account deficits sustainable?

Summarizing our results so far, the foreign sector does not seem to have worked as a

constraint on the growth of the Spanish economy over the long run, namely, throughout

the periods 1850-1913 and 1940-2000. If any, this might have occurred in some shorter

and specific subperiods. In particular, we have identified two such subperiods: 1940-

1959, i.e., the “autarky” period following the end of the Spanish Civil War, which led to

a situation of virtual bankruptcy of the economy; and 1959-1974, i.e., the episode of

highest economic growth in Spanish economic history.

Overall, these results would roughly support the arguments put forward by the

intertemporal approach to the current account so that, far from being a constraint on

further growth, the external deficit would have allowed the Spanish economy to smooth

her level of aggregate consumption over the long run. However, we also mentioned in

section 2 that such an “optimistic” outcome requires that the external deficit were

sustainable over the long run. In the rest of this section, we will analyze this issue, that

is, whether the external deficit was sustainable throughout 1850-1913 and 1940-2000.

To this end, we will follow the approach utilized in Bajo-Rubio, Díaz-Roldán and

Esteve (2010) for the analysis of the (analogous) case of the government deficit.

The sustainability of external deficits is related to the issue of long-run solvency.

A current account deficit is regarded as sustainable when, if maintained in the indefinite

future, it does not violate the nation’s solvency constraint; and a nation is said to be

solvent if the present-value budget constraint, i.e., its intertemporal budget constraint

(IBC) holds. In other words, a deficit can be sustainable if the country can borrow.

However, if the interest rate on the external debt exceeds the growth rate of the

economy, debt dynamics would lead to an ever-increasing ratio of debt to GDP. The

dynamics of debt accumulation could be stopped only if the ratio of the external deficit

to GDP would turn to be a surplus.

The customary framework used to test for the sustainability of external deficits

follows from the intertemporal approach to the current account, and starts from the

nation’s IBC, written in terms of GDP shares:

18

1

1

10

1

11lim

11

jtt

j

jjttj

j

t nfaErgnxE

rgnfa

where nfa and nx denote, respectively, net foreign assets and net exports of goods and

services, both as ratios to GDP; E is the expectations operator; and g and r stand,

respectively, for the rate of growth of real GDP and the real interest rate, both assumed

to be constant for simplicity. The condition for current account sustainability is:

i.e., the transversality condition; or, equivalently:

By multiplying both sides of this equation by −1, so that the country is a net debtor, we

can see that solvency requires that the country must run expected future surpluses equal,

in present-value terms, to the current value of its outstanding net liabilities vis-à-vis the

rest of the world.

The cointegration framework to test for the IBC follows once first differences

are taken in the above expression for the nation’s IBC:

Δ nfat=−∑j=0

∞

(1+g1+r )

j+1Et Δnx t+ j+1+ lim

j→∞ (1+g1+r )

j+1Et Δnfat+ j+1

(2)

so that sustainability would require:

limj→∞( 1+g

1+r )j+1

E t Δnfat+ j+1=0 (3)

Under a no-Ponzi scheme rule, the right-hand side of equation (2) will be

stationary as long as net exports and the stock of net foreign assets are all stationary in

first differences. In order to test for condition (3), the usual procedure consists of testing

for the stationarity of

∆nfat = cabt = expt − impt

where expt and impt denote, respectively, the GDP ratios of the exports of goods and

services, and the imports of goods and services plus net interest payments and net

transfer payments; that is, expt − impt would equal the current account balance, cabt, as a

ratio to GDP. Provided that both expt and impt are I(1) with a cointegration relationship

(1, −1), then one should test for the linear restriction β=1 in a regression model of the

form:

expt = α + βimpt + εt (4)

19

011lim 1

1

jtt

j

jnfaE

rg

10

1

11

jttj

j

t nxErgnfa

where εt is an error term. In particular, translating Quintos’s (1995) definitions to the

case of the current account deficit, if expt and impt are cointegrated and β = 1, the

external deficit would be strongly sustainable; whereas, if 0 < β < 1, the deficit would

be only weakly sustainable. Finally, if β = 0 the deficit would be unsustainable.

The data sources for expt, impt and cabt are Prados de la Escosura (2010) for

1850-1913 and Tena (2007) for 1940-2000, expressed as GDP shares using the GDP

series of Prados de la Escosura (2003). Specifically, for 1850-1913 expt, impt and cabt

come from Appendix 1 in Prados de la Escosura (2010): columns 1 plus 2; 5 plus 6

minus 10 minus 11; and 12, respectively. In turn, for 1940-2000 expt, impt and cabt

come from Table 8 in Tena (2007): series 29 plus 33; 45 plus 49 minus 70 minus 73;

and 74, respectively.

We begin by testing for the order of integration of the variables expt, impt and

cabt using the tests of Ng and Perron (2001). These authors proposed using the tests

statistics MZα and MZt , which are modified versions of the Zα and Z t Phillips-Perron

tests aimed to improve the tests with regard to both size distortions and power. The

results are shown in Table 4, and the null hypothesis of no stationarity cannot be

rejected, independently of the test and the period, for expt and impt; in other words, these

two variables would be integrated of order one, so that we can proceed to estimate a

cointegration equation such as (4). In turn, the null hypothesis of no stationarity is

rejected according to both tests, at the 5% level in 1850-1913 and at the 1% level in

1940-2000, for cabt; in other words, the current account balance (as a ratio to GDP)

would be a stationary variable. This latter result can be interpreted as an indirect test of

sustainability, so that a stationary current account balance can be deemed as sustainable.

Notice, by the way, that the simple rule mentioned in section 2, according to which a

current account deficit in excess of 5% of GDP was regarded as unsustainable (Freund,

2005) does not apply in our case. Indeed, the maximum levels attained by the current

account deficit-GDP ratio were 4.79% in 1858 and 4.06% in 1976, for the 1850-1913

and 1940-2000 periods, respectively.

[Table 4 here]

20

A more formal test on the sustainability of the current account deficit involves

the estimation of equation (4); the results appear in Table 5. The equation has been

estimated using the method of Phillips and Hansen (1990), and we have also tested for

the presence of a time trend in the long-run relationship, which allows us to test for the

presence of stochastic or deterministic cointegration, i.e., when the trend is included or

not in the estimated long-run relationship4. As in Table 1, the figures in parentheses are

Wald test statistics on the null hypothesis that the estimated coefficients are equal to

zero. We also include in the last three lines of the table the coefficient of determination,

R2; the cointegration test proposed by Phillips and Ouliaris (1990), CR { Z t¿ ; and a Wald

test on the null hypothesis that the estimate of the coefficient β is equal to one.

[Table 5 here]

Beginning with the period 1850-1913, the trend is significantly different from

zero, so that we will focus our comments on column one5. The estimate of β is 0.44,

positive and significant, but significantly different from one, according to the Wald test

in the last row of the table. Finally, the Phillips-Ouliaris test allows us to reject the null

hypothesis of no cointegration, so in this case there would be stochastic cointegration.

As a result, the current account deficit was weakly sustainable during this period.

In turn, for the period 1940-2000, the trend is not significantly different from

zero, so that our preferred specification will be that in column four. The estimate of β is

0.94, positive and significant, but in this case not significantly different from one.

Again, the Phillips-Ouliaris test detects the presence of cointegration, which now would

be deterministic cointegration. Consequently, during the second period the current

account deficit was strongly sustainable.

4 The concepts of stochastic and deterministic cointegration were introduced by Ogaki and Park (1997).

5 Notice, on the other hand, that the estimate of β in column two, 1.52, is implausibly high because a β greater than one would not be consistent with a current account deficit, since expt

would be growing at a faster rate than impt.

21

5. Conclusions

The Spanish economy has experienced a steady process of growth through the period

1850-2000, following a similar evolution to that of the rest of Western Europe.

However, only in the second half of the 20th century Spain could catch-up with these

countries, offsetting, at least partially, her relative retardation dating back to the start of

industrialization. Although the periods of higher growth have been related to increases

in the degree of external openness, some concerns have been also raised about whether

a more open foreign sector could become a constraint on further growth via

unsustainable trade deficits. This argument would be justified given the structural

problems underwent by the Spanish economy, as shown in the weakness of exports, and

the need for some essential imports.

In this paper, we have analyzed whether or not the balance of payments would

have meant an impediment to further GDP growth vis-à-vis Western Europe, over the

period 1850-2000. More specifically, we have examined the two subperiods 1850-1913

and 1940-2000, on the grounds of availability of data on the current account balance; so,

leaving aside the interwar years, characterized by the disruption of the world economic

order related to the first globalization. To this end, we have followed a simple approach,

calculating the so-called balance of payments-constrained growth rate from estimated

income and real exchange rate elasticities for exports and imports, and allowing for the

role of capital inflows. Then, this balance of payments-constrained growth rate is

compared with the actual GDP growth rate.

We found that for both 1850-1913 and 1940-2000, the Spanish rate of GDP

growth was only slightly above the balance of payments-constrained growth rate, so that

the foreign sector does not seem to have worked as a constraint on the growth of the

Spanish economy over the long run, i.e., along the two subperiods 1850-1913 and 1940-

2000.

Next, in order to have a more precise picture, and given the potentially different

behaviour of the series within every period, we have repeated the above exercise across

a set of different subperiods. In the case of 1850-1913, we have analyzed the two

subperiods 1850-1891 and 1892-1913, allowing for a possible break point in 1891 that

relates to the inward-looking policy stance adopted at the end of the 19th century.

22

Regarding 1940-2000, several alternative subperiods have been examined: (i) 1951-

2000, 1951-1974 and 1975-2000, attending to the evolution of the GDP series; (ii)

1940-1959 and 1960-2000, attending to the evolution of the foreign trade series; and

(iii) 1960-1974 and 1986-2000, related to the end of the expansionary phase in 1974,

and the Spanish integration into the EU in 1986, respectively. As an additional check of

the robustness of our results, we also tried two alternative proxies for foreign output. In

general, the results did not change substantially across subperiods, and depending on

how foreign output was measured. We have detected just two subperiods for which the

actual growth rate exceeded to a certain extent the balance-of-payments constrained

growth rate. They were: (i) 1940-1959, i.e., the “autarky” period that ended into a

situation of virtual bankruptcy of the economy (with an excess of between 1.3 and 2.6

points, according to the particular proxy used for foreign output); and (ii) 1959-1974,

i.e., the episode of highest economic growth in Spanish economic history (with an

excess of around 2 points).

Finally, since the above results tend to suggest that the external deficit would

have allowed the Spanish economy smooth her level of aggregate consumption over the

long run, we have tested whether the external deficit was sustainable throughout the

periods 1850-1913 and 1940-2000. We made this by testing the country’s intertemporal

solvency, i.e., the fulfilment of the nation’s IBC, by means of the estimation of a long-

run relationship between exports and imports of goods and services as ratios to GDP.

For both periods, we found that the two variables were cointegrated, and the estimated

regression coefficient was positive and significant. Since this coefficient was lower than

one for 1850-1913, and not significantly different from one for 1940-2000, the current

account deficit would prove to be weakly sustainable and strongly sustainable during

the first and the second period, respectively.

Our results for the first period, 1850-1913 would agree with those of Prados de

la Escosura (2010), who found no evidence on an external constraint on growth during

that time; and argues that, by making possible the arrival of foreign capital inflows,

growth would have proved to be higher in periods where current account deficits

predominated. On the other hand, for the second period, 1940-2000, some evidence was

found on the external sector as constraint on growth for the two subperiods 1940-1959

23

and 1959-1974, i.e., for the whole Franco regime. While the timid reforms of the 1950s

would appear as a precondition for the successful performance of the 1960s (Prados de

la Escosura, Rosés and Sanz-Villarroya, 2011), the later reforms associated with the

Stabilization Plan of 1959 and subsequent measures, seem to have been insufficient to

allow for a sustained growth, fully free of restrictions originated into the balance of

payments. Only in more recent times (and, especially, following integration into the

EU) income elasticities for exports and imports seem to have reached some more

favourable values, according to Krugman’s (1989) terminology. In other words, the

increased external openness, coupled with the structural transformations operated in the

Spanish foreign trade over the last thirty years, seem to have faded out the role of the

external constraint on growth.

Summarizing, the foreign sector does not seem to have worked as a constraint on

the growth of the Spanish economy over the long run, that is, throughout the periods

1850-1913 and 1940-2000. If any, this might have occurred in some shorter and specific

subperiods, such as 1940-1959 and 1959-1974. Overall, these results would roughly

support the arguments put forward by the intertemporal approach to the current account

so that, far from being a constraint on further growth, the external deficit would have

allowed the Spanish economy smooth her level of aggregate consumption over the long

run. This outcome would hold given the finding of sustainability of the external deficit

over the long run.

To conclude, recall that, once Spain joined the Economic and Monetary Union

(EMU) of the EU after 1999, the disappearance of the exchange rate risk made easier

borrowing in international markets, due to both an increase in the supply of funds

available and a decrease in its cost. In other words, the allowable external deficit would

be higher in a monetary union (Blanchard and Giavazzi, 2002). Even though such an

argument certainly softens the practical significance of the external constraint on growth

for a country belonging to a monetary union, this should not mean neglecting the size of

the external deficit, however. For instance, a large external deficit could reflect an

unbalanced growth pattern and a loss of competitiveness of the economy, leading to a

higher inflation than in the rest of the union, and hence to a real exchange rate

appreciation. However, in a monetary union is no longer possible to offset such a loss of

competitiveness by means of the depreciation of the nominal exchange rate, requiring

24

instead a lower growth of domestic prices that would affect negatively future growth

prospects. In other words, in a monetary union, the external constraint would be still

binding, and the financing of current account imbalances would not be without limit6.

The non clear-cut results on current account sustainability in the Spanish case for the

period 1970-2007, found in Bajo-Rubio, Díaz-Roldán and Esteve (2011) can be

interpreted in this light.

6 The role of current account imbalances within a monetary union is discussed at length in Catte (1998); see also L’Hotellerie-Fallois and Peñalosa (2006) for a particular discussion of the Spanish case within EMU.

25

ReferencesAghion, P. and Howitt, P. (1998): Endogenous Growth Theory, The MIT Press, Cambridge, MA.

Alonso, J. A. (1999): “Growth and the external constraint: Lessons from the Spanish case”, Applied Economics 31, 245-253.

Bai, J. and Perron, P. (1998): “Estimating and testing linear models with multiple structural changes”, Econometrica 66, 47-78.

Bajo-Rubio, O. and Torres, A. (1992): “El comercio exterior y la inversión extranjera directa tras la integración de España en la CE (1986-90)”, in Viñals, J. (ed.): La economía española ante el Mercado Unico europeo. Las claves del proceso de integración, Alianza, Madrid, 167-228.

Bajo-Rubio, O., Díaz-Roldán, C. and Esteve, V. (2010): “On the sustainability of government deficits: Some long-term evidence for Spain, 1850-2000”, Journal of Applied Economics 13, 263-281.

Bajo-Rubio, O., Díaz-Roldán, C. and Esteve, V. (2011): “Sustainability of external imbalances in the OECD countries”, Working Paper on International Economics and Finance 11-07, Spanish Association of International Economics and Finance.

Blanchard, O. and Giavazzi, F. (2002): “Current account deficits in the euro area: The end of the Feldstein-Horioka puzzle?”, Brooking Papers on Economic Activity 2, 147-186.

Carreras, A. and Tafunell, X. (2004): “The European Union economic growth experience, 1830-2000”, in Heikkinen, S. and Van Zanden, J. L. (eds.): Explorations in economic growth: Essays in measurement and analysis, Aksant, Amsterdam, 63-87.

Carreras, A. and Tafunell, X. (2005): “El crecimiento económico en la Unión Europea, 1830-2000”, Cuadernos Económicos de ICE 70, 25-50.

Catte, P. (1998): “Current accounts: Are they still relevant for and within a monetary union?”, in Oesterreichische Nationalbank: Current account imbalances in East and West: Do they matter?, Vienna, 144-168.

Corden, W. M. (2007): “Those current account imbalances: A sceptical view”, The World Economy 30, 363-382.

Crafts, N. (1992): “Productivity growth reconsidered”, Economic Policy 15, 388-426.

De la Dehesa, G., Ruiz, J. J. and Torres, A. (1991): “Spain”, in Papageorgiou, D., Michaely, M. and Choksi, A. M. (eds.): Liberalizing foreign trade: The experience of New Zealand, Spain, and Turkey, Basil Blackwell, Oxford, 137-262.

Donges, J. B. (1971): “From an autarchic towards a cautiously outward-looking industrialization policy: The case of Spain”, Weltwirtschaftliches Archiv 107, 33-75.

26

Donges, J. B. (1976): La industrialización en España: Políticas, logros, perspectivas, Oikos-Tau, Barcelona.

Dumke, R. H. (1990): “Reassessing the Wirtschaftswunder: Reconstruction and postwar growth in West Germany in an international context”, Oxford Bulletin of Economics and Statistics 52, 451-491.

Edwards, S. (1992): “Trade orientation, distortions and growth in developing countries”, Journal of Development Economics 39, 31-57.

Edwards, S. (1993): “Openness, trade liberalization, and growth in developing countries”, Journal of Economic Literature 31, 1358-1393.

Edwards, S. (2002): “Does the current account matter?”, in Edwards, S. and Frankel, J. A. (eds.): Preventing currency crises in emerging markets, The University of Chicago Press, Chicago, 21-69.

Fanjul, O. and Segura, J. (1977): Dependencia productiva y exterior de la economía española 1962-70, Fundación del INI, Madrid.

Freund, C. (2005): “Current account adjustment in industrial countries”, Journal of International Money and Finance 24, 1278-1298.

Grossman, G. M. and Helpman, E. (1994): “Endogenous innovation in the theory of growth”, Journal of Economic Perspectives 8, 23-44.

Goldstein, M. and Khan, M. S. (1985): “Income and price effects in foreign trade”, in Jones, R. W. and Kenen, P. B. (eds.): Handbook of International Economics, vol. 2, North-Holland, Amsterdam, 1041-1105.

Harrod, R. (1933): International Economics, Cambridge University Press, Cambridge.

Herranz-Loncán, A. and Tirado-Fabregat, D. A. (1996): “La restricción exterior al crecimiento económico español (1870-1913)”, Revista de Historia Económica 14, 11-49.

Krugman, P. (1989): “Differences in income elasticities and trends in real exchange rates”, European Economic Review 33, 1031-1054.

L’Hotellerie-Fallois, P. and Peñalosa, J. (2006): “El diagnóstico del déficit exterior español dentro de la UEM”, Cuadernos de Información Económica 192, 17-29.

Maddison, A. (2010): Historical Statistics of the World Economy: 1-2008 AD, available at http://www.ggdc.net/MADDISON/oriindex.htm.

McCombie, J. S. L. and Thirlwall, A. P. (2004): Essays on Balance of Payments Constrained Growth: Theory and Evidence, Routledge, London.

Milesi-Ferretti, G. M. and Razin, A. (1996): Current-account sustainability, Princeton Studies in International Finance 81, Princeton University.

27

Molinas, C. and Prados de la Escosura, L. (1989): “Was Spain different? Spanish historical backwardness revisited”, Explorations in Economic History 26, 385-402.

Nadal, J. (1975): El fracaso de la revolución industrial en España, 1814-1913, Ariel, Barcelona.

Ng, S. and Perron, P. (2001): “Lag length selection and the construction of unit root tests with good size and power”, Econometrica 69, 1529-1554.

Obstfeld, M. (1982): “Aggregate spending and the terms of trade: Is there a Laursen-Metzler effect?”, Quarterly Journal of Economics 97, 251-270.

Obstfeld, M. and Rogoff, K. (1995): “The intertemporal approach to the current account”, in Grossman, G. and Rogoff, K. (eds.): Handbook of International Economics (Vol. 3), North-Holland, Amsterdam, 1731-1799.

Ogaki, M. and Park, J. Y. (1997): “A cointegration approach to estimating preference parameters”, Journal of Econometrics 82, 107-134.

Phillips, P. C. B. and Hansen, B. E. (1990): “Statistical inference in instrumental variables regression with I(1) processes”, Review of Economic Studies 57, 99-125.

Phillips, P. C. B. and Ouliaris, S. (1990): “Asymptotic properties of residual based tests for cointegration”, Econometrica 58, 165-193.

Prados de la Escosura, L. (1988): De imperio a nación. Crecimiento y atraso económico en España (1780-1930), Alianza Editorial, Madrid.

Prados de la Escosura, L. (2003): El progreso económico de España (1850-2000), Fundación BBVA, Bilbao.

Prados de la Escosura, L. (2007): “Growth and structural change in Spain, 1850-2000: A European perspective”, Revista de Historia Económica/Journal of Iberian and Latin American Economic History 25, 147-181.

Prados de la Escosura, L. (2010): “Spain’s international position, 1850-1913”, Revista de Historia Económica/Journal of Iberian and Latin American Economic History 28, 173-215.

Prados de la Escosura, L. and Rosés, J. R. (2009): “The sources of long-run growth in Spain, 1850-2000”, Journal of Economic History 69, 1063-1091.

Prados de la Escosura, L. and Sanz, J. C. (1996): “Growth and macroeconomic performance in Spain, 1939-93”, in Crafts, N. and Toniolo, G. (eds.): Economic Growth in Europe since 1945, Cambridge University Press, Cambridge, 355-387.

Prados de la Escosura, L., Rosés, J. R. and Sanz-Villarroya, I. (2011): “Economic reforms and growth in Franco’s Spain”, Revista de Historia Económica/Journal of Iberian and Latin American Economic History 29, forthcoming.

28

Quintos, C. E. (1995): “Sustainability of the deficit process with structural shifts”, Journal of Business and Economic Statistics 13, 409-417.

Razin, A. (1995): “The dynamic-optimizing approach to the current account: Theory and evidence”, in Kenen, P. B. (ed.): Understanding interdependence: The macroeconomics of the open economy, Princeton University Press, Princeton, 169-198.

Rodríguez, F. and Rodrik, D. (2001): “Trade policy and economic growth: A skeptic’s guide to the cross-national evidence”, in Bernanke, B. S. and Rogoff, K. (eds.): NBER Macroeconomics Annual 2000, 261-325.

Sachs, J. (1981): “The current account and macroeconomic adjustment in the 1970’s”, Brookings Papers on Economic Activity 1, 201-268.

Sachs, J. (1982): “The current account in the macroeconomic adjustment process”, Scandinavian Journal of Economics 84, 147-159.

Sardà, J. (1948): La política monetaria y las fluctuaciones de la economía española en el siglo XIX, Ariel, Barcelona. 

Sardà, J. (1970): “El Banco de España (1931-1962)”, in El Banco de España: Una historia económica, Banco de España, Madrid, 419-479. 

Schumpeter, J. A. (1934): The theory of economic development, Harvard University Press, Cambridge, MA.

Segura, J. and García-Viñuela, E. (1978): “El déficit comercial de la economía española”, Información Comercial Española 536, 13-23.

Serrano-Sanz, J. M. (1997): “Sector exterior y desarrollo en la economía española contemporánea”, Papeles de Economía Española 73, 308-335.

Serrano-Sanz, J. M., Sabaté-Sort, M. and Gadea-Rivas, M. D. (2008): “Una mirada ingenua sobre las series del sector exterior, 1869-1999”, Revista de Historia Económica/Journal of Iberian and Latin American Economic History 26, 83-108.

Svensson, L. E. O. and Razin, A. (1983): “The terms of trade and the current account: The Harberger-Laursen-Metzler effect”, Journal of Political Economy 91, 97-125.

Taylor, A. M. (2002): “A century of current account dynamics”, Journal of International Money and Finance 21, 725-748.

Tena, A. (1995): “The Spanish foreign sector 1885-1985”, in Martín-Aceña, P. and Simpson, J. (eds.): The economic development of Spain since 1870, Edward Elgar, Aldershot, 402-418.

Tena, A. (2005): “Sector exterior”, in Carreras, A. and Tafunell, X. (eds.): Estadísticas históricas de España: Siglos XIX-XX (2nd edition), Fundación BBVA, Bilbao, 573-644.

29

Tena, A. (2006): “Spanish protectionism during the Restauración, 1875-1930”, in Dormois, J.-P. and Lains, P. (eds.): Classical Trade Protectionism 1815-1914, Routledge, London, 265-297.

Tena, A. (2007): “New series of the Spanish foreign sector, 1850-2000”, Working Paper in Economic History 07-14, Departamento de Historia Económica e Instituciones, Universidad Carlos III de Madrid.

Thirlwall, A. P. (1979): “The balance of payments constraint as an explanation of international growth rate differences”, Banca Nazionale del Lavoro Quarterly Review 128, 45-53.

Thirlwall, A. P. (2011): “Balance of payments constrained growth models: History and overview”, Discussion Paper 11/11, School of Economics, University of Kent.

Thirlwall, A. P. and Hussain, M. N. (1982): “The balance of payments constraint, capital flows and growth rate differences between developing countries”, Oxford Economic Papers 34, 498-510.

Tortella, G. (2000): The Development of Modern Spain: An Economic History of the Nineteenth and Twentieth Centuries, Harvard University Press, Cambridge, MA.

30

Table 1. Estimated exports and imports elasticities, 1850-1913 and 1940-2000

ε X ,Y ¿ ε M , Y ε X ,Q ε M , Q

1850-1913 1.70a

(304.4) 1.93a

(202.1) 1.51a

(16.98) 0.30(0.60)

1850-1891 2.30a

(139.7) 2.69a

(66.60) 0.41 (1.09)

−0.89 (2.13)

1892-1913 0.73a

(16.20) 1.43a

(34.31) 0.67 (1.52)

−0.35 (0.44)

1940-2000 2.60a

(158.3) 1.79a

(639.9) 0.13 (0.12)

0.32 (3.25)

1951-2000 2.79a

(219.8) 1.85a

(944.6)−0.65 (1.53)

1.01a

(16.85)

1951-1974 1.73a

(73.54) 1.57a

(1372.1)−1.87a

(10.61)−0.07 (0.17)

1975-2000 4.18a

(224.3) 2.47a

(593.8)−0.35 (1.18)

0.83a

(24.94)

1940-1959 0.71a

(28.44) 1.33a

(110.4)−0.60a

(66.47)−0.30a

(23.69)

1960-2000 3.65a

(874.3) 2.10a

(1680.9) 0.17 (0.49)

1.34a

(73.79)

1960-1974 2.54a

(106.2) 1.75a

(1517.6)−0.42 (0.64)

0.37b

(5.79)

1986-2000 4.59a

(49.76) 3.09a

(223.4) 0.08 (0.01)

0.25 (0.36)

Note: a and b denote significance at the 1% and 5% levels, respectively. The critical values for the Wald tests (distributed as a χ2 with one degree of freedom) are 6.63 and 3.84 (at the 1% and 5% levels).

Source: Own calculations; see the text.

31

Table 2. Actual and balance of payments-constrained growth rates, 1850-1913 and 1940-2000

Y Y ¿ Y B

1850-1913 1.52 2.00 1.381850-1891 1.47 1.83 0.94

1892-1913 1.28 2.29 0.72

1940-2000 4.17 2.99 3.65

1951-2000 4.61 3.30 7.43

1951-1974 6.23 4.45 6.13

1975-2000 3.02 2.25 3.78

1940-1959 3.00 2.73 1.68

1960-2000 4.71 2.97 5.72

1960-1974 7.49 4.23 5.67

1986-2000 3.39 2.19 4.13

Source: col. (1), Prados de la Escosura (2003); col. (2), Carreras and Tafunell (2004, 2005); col. (3), own calculations. See the text.

32

Table 3. Actual and balance of payments-constrained growth rates, 1870-1913 and 1940-2000. Alternative data for foreign output

Western Europe’s GDP Weighted average of 6 trading partners’ GDP

Y Y ¿ Y B Y Y ¿ Y B

1870-1913 1.79 2.09 2.06 1.79 1.47 1.991870-1891 1.99 1.75 2.02 1.99 1.23 2.21

1892-1913 1.28 2.41 0.74 1.28 1.69 0.75

1940-2000 4.17 2.76 3.50 4.17 2.04 2.95

1951-2000 4.61 3.18 7.45 4.61 2.83 7.45

1951-1974 6.23 4.30 6.11 6.23 3.82 6.21

1975-2000 3.02 2.20 3.80 3.02 1.96 3.73

1940-1959 3.00 2.30 1.36 3.00 0.97 0.34

1960-2000 4.71 2.82 5.75 4.71 2.39 5.49

1960-1974 7.49 3.97 5.64 7.49 3.19 5.23

1986-2000 3.39 2.12 4.15 3.39 1.43 3.96

Source: col. (1) and (4), Prados de la Escosura (2003); col. (2) and (5), Maddison (2010); col. (3) and (6), own calculations. See the text.

33

Table 4. Ng-Perron tests for unit roots

1850-1913

MZα MZ t

expt

impt

cabt

−0.34−3.46

−8.47b

−0.19−1.13

−2.00b

1940-2000

MZα MZt

expt

impt

cabt

3.36 3.78

−159.8a

2.93 2.80

−8.90a

Note: a and b denote significance at the 1% and 5% levels, respectively. The critical values (taken from

Ng and Perron, 2001) are −13.8 and −8.1 (at the 1% and 5% levels) forMZα ; and −2.58 and

−1.98 (at the 1% and 5% levels) forMZt .Source: Own calculations; see the text.

Table 5. Estimation of a long-run relationship between exports and imports of goods and services

1850-1913 1940-2000

α

β

trend

4.38a

(8.96)

0.44a

(7.15)

0.12a

(46.43)

−5.08a

(9.59)

1.52a

(69.95)

−

1.39 (0.54)

0.83a

(33.51)

0.04 (0.42)

0.12(0.05)

0.94a

(648.4)

−

R2

CR { Z t¿

W: β=1

0.87

−4.36a

11.43a

0.65

−4.20a

8.28a

0.96

−3.89b

1.33

0.95

−3.97a

2.99

Note: a and b denote significance at the 1% and 5% levels, respectively. The critical values for the Wald tests (distributed as a χ2 with one degree of freedom) are 6.63 and 3.84 (at the

1% and 5% levels). The critical values for CR { Z t¿ (taken from Phillips and Ouliaris, 1990) are −4.30 and −3.74 (at the 1% and 5% levels) for col. 1 and 3; and −3.84 and −3.27 (at the 1% and 5% levels) for col. 2 and 4.

Source: Own calculations; see the text.

34