comparative perspective between Latin america and east Europe.

lEEJ LIBERALIZATION and its Consequences

2. Economic performance and the state in Latin America Victor Bulmer-Thomas

This chapter begins with an examination of economic ,performance in Latin America with an emphasis on the 20th century. This is a long enough period to establish which countries have been successes or failures according to well­ established criteria. The data set used also allows for determination of tbe sub-periods in which success or failure Itas been most marked. The regional variations in Latin America's economic performance have many possible expla­ nations. One of these is the role of the state. In the second part of the chapter I outline the theory of the state with reference to economic development, drawing anention to the different ways the state can influence the allocation of resources and affecttbe rate of economic growth. The state will have an impact on both the supply- and demand-side of the economy. In the third part of the chapter I look at the supply-side, focusing in particular on the markets for factors of production (land, labor and capital) and on technical progress. In the next part of the chapter I explore the impact of the state on the demand-side and I focus on both macro- and microeconomic policies. In the final section of the chapter, I draw out the main conclusions on the role of the state in Latin America using a chronological approach. This starts with export-led growth up to the 1920s, then moves to the inward-looking phase of development that began in the 1930s and ends with the neo-liberal experiment currently in progress.

ECONOMIC PERFORMANCE IN LATIN AMERICA SINCE 1900

At the begirming of tbe 20th century, after nearly a century of independence, few Latin American countries had achieved sustainable long-run economic development and none had acttieved the high expectations vested in U,e region at the birth of the republics. Although cross-country statistics are woefully inadequate, a clear picture emerges for 1900 in which Argentina and Uruguay can be classified as 'successful', Cttile and Cuba as 'moderately successful' and Mexico as a special case. The other Latin American countries were all failures as far as economic development was concerned.

12

13Economic performance and the stale in Latin America

There are several reasons for this judgment. First, of the nine countries for which we have figures on Gross Domestic Product (GOP) per head in 1900 (see Table 2.1), only four - Argentina , Chile. Cuba and Mexico - had achieved even ten percent of the comparable figure for the United States. There are no figures for Uruguay , but when GOP per head was calculated for the first time in 1940 it was the highest in Latin America , making it highly probable that Uruguay was among the leaders in 1900. I Secondly , illiteracy rates, for which we have data for all countries except the Dominican Republic and Nicaragua, confirm the superior stalus of Argentina, Chile, Cuba and Uruguay (see Table 2.1). Mexico, however, lags behind in 13th place, suggesting that the social development in the Porftriat02 fell far behind the econO/nic achievements; that is why I bave labeled Mexico in 1900 a special case .

The third statistic we can use is exports per head . ~uring a period of export­ led growth , such as Latin Arnerica experienced before World War I, exports per head are highly correlated wilh GOP per head. While the figures for 1900 are incomplete, we do have figures for c. 1912 (see Table 2 .1) and the ranking of countries is unlikely 10 have changed by much in the intervening years. Once again, Argentina, Chile, Cuba, Mexico and Uruguay emerge as the success stories. The fourtb statistic is bank deposits per bead . Financial intermediation is a good proxy for the level of development and its depth can be measured by this indicator. As Table 2.1 shows, Argentina stands out, with Chile and Uruguay following at some distance. The other countries for which data exisl (for example Venezuela) appear to have had very shallow finance at the beginning of the century .

Argentina is the most straightforward case of a successful country if judged by the statistics in Table 2.1. It ranks first in GOP and bank deposits per head, second in exports per head and literacy . Uruguay ranks ft.rst in hteracy, second in bank deposits and third in exports per head (there are no figures for GOP per bead), but already by 1900 Uruguay had shifted from an export-led growth model to one in which tbe internal market was probably growing as fast as exports 3 Chile and Cuba (still not fully independent in 1900)4 perform well in relation to the rest of Latin America, but lag some way behind Argentina on all four criteria. Mexico remains a special case, in which economic development had effectively been postponed by 50 years after independence in the 1820s as a result of the cycle of civil wars and foreign intervention.

The Latin America panorama is, therefore, a gloomy one at the start of the 20th century. Of the 15 failures, 5 some were quite spectacular. Brazil in 1900 had a GDP per head only 4.8 percent of the US level despite the advances in industrialization since 1870. Many republics had exports per head valued at less than ten dollars while bank deposits per head outside of the southern cone were uniformly 'low. Comparisons of GOP per head require the use of exchange rates and it might be argued Ihat these bias the results against Latin Arnerica.

14 Liberalization and its consequences

Table 2.1 Economic and social indicators/or Latin America: c.1900

GDP per bead illiteracy rate Exports per BanIc deposits Country as % of US (%)' head ($)b per head($)b

Argentina' 29.7 Bolivia Brazil 4.8 Chiled 19.1 Colombia 8.0 Costa Rica Cubad 18.4 Dom.Rep. Ecuador 6.0 El Salvador Guatemala Haiti Honduras Mexicoe 17.7 Nicaragua Panama Paraguay Peru 7.0 Uruguay' Venezuela 7.2

Latin America 12.5f

48.7 81.5 65.3 56.5 66.0 64.4 54.0

66.9 73.7 88.\ 92.0 71.7 75.6

82.7 68.6 75.7 40.6 72.2

66.1f

62.0 18.6 14.2 44.7

6.4 27.1 64.7 15.5 7.9 8.3 7.2 6.1 4.7

10.7 10.8

8.6 9.4

50.3 10.5

20.4

75.7 3.3 9.4

26.0

1.6 2.3 0.9

0.9 29.5

1.2

NOtes: a Percent of population over I5 . b c. 1912. c 'Success' story. d 'Moderate success' stOry. e Special case. r Six co untries only (Argentina, Brazil, Chile, Colombia, Mexico, and Venezuela).

Sources: Bulmer·Thomas (1995); Thorp (1998).

However, the exchange rates used in Table 2.1 for GDP per head are purchasing-power parity exchange rates and not official exchange rates. Using the laner, Brazil's GDP per head would have been only four percent of the US figure in 1900 - not 4.8 percent.6

Throughout the 20th century (see Table 2.2), GDP per head for Latin America as a whole bas remained the same proportion of US GDP per head. This suggests a simple definition of success or failure for 20th century economic

Economic performance and the Slate in LAtin America 15

performance. Those countries that closed the gap on the United States can be defined as successes; where the gap widened, the country can be classified as a failure. This analysis can be applied both across the whole century and to different sub-periods. It has the additional advantage tbat ' success' in this context means that the country outperformed the Latin American average while 'failure' means that it underperfonned.

Looking at the century as a whole, only three countries can be clearly classified as success stories: Brazil, Colombia and Venezuela (see Table 2.2). Of these three countries, Brazil is the clearest case; its GDP per head reached a higher proportion of the US figure in every decade until the 1980s. Indeed, without tbe successful economic perfonnance in Brazil, the Latin American perfonnance relative to the United States would have looked truly dismal for most of the 20th century.

Table 2.2 GDP per head as percentage of US GDP per head: 1900-95

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 1995

AIgentina 29.7 32.4 27.0 26.0 26.0 23.4 22.2 23 .1 21.9 15.5 18.1 Bolivia 7.9 5.6 5.7 5.6 3.9 4.0 Brazil 4.8 4.8 5.1 5.9 6.4 6.5 8.4 8.7 12.3 10.7 10.4 Chile 19.1 19.8 16.6 23.3 19.4 17.5 17.7 16.5 15.2 14.9 18.0 Colombia 8.0 8.5 9.0 10.7 11.7 10.9 10.9 10.4 10.7 10.2 11.1 Costa Rica 14.6 12.9 13.1 11.2 12.2 12.7 14.0 10.9 11.4 Cuba 18.4 24.0 21.1 17.0 15.1 11.5 10.1 7.2 10.3 9.3 6.2 Dom. Rep. 7.4 7.8 7.4 8.6 6.9 7.0 Ecuador 6.0 6.3 6.9 7.2 6.4 7.0 7.4 6.9 8.6 7.0 7.1 E1 Salvador 8.6 8.3 7.6 8.3 8.6 7.9 6.5 4.8 5.5 Guatemala 12.4 11.4 15.4 9.4 8.8 8. 1 8.2 6.1 6. 1 Haiti 3.9 3.1 2.3 2.5 1.6 1.1 Honduras 11.4 12.3 7.9 6.9 6.2 5.4 4,9 3.7 3.8 Mexico 17.7 18.4 18.2 14.6 14.4 13.9 15.9 17.1 18.5 15.0 14.1 Nicaragua 9.1 9.3 7.4 6.6 7.5 8.3 5,0 2.6 2.3 Panama 13.9 14.6 17.3 17.4 12.8 14.2 Paraguay 12.8 8.9 7.9 7,0 9.8 7.6 7.2 Peru 7.0 8.7 10.1 12.6 12.4 1t.2 12.6 11.9 11.1 6.7 7.3 Uruguay 26.7 26.2 23.8 18 .8 18.3 15,7 17.5 Venezuela 7.2 6.7 7.5 19.0 20.2 29.5 29.3 25.8 24.3 16.9 16.1 Latin Americaft 12.5 13.3 12.4 12.9 12.9 12.5 13.6 13.7 15.4 12.7 12.8 Latin Americab 11.9 12.7 12.6 14.0 11.3 11.4

Notes: a Six countries only (Argentina, Brazil, Chile, Colombia., Mexico and Venezuela) . b Al120 counlries.

Source: Derived from Thorp (1998).

16 Liberalization and lIS consequences

The Colombian and Venezuelan stories are somewhat differen!. Both countries did indeed close the gap, but in the case of Colombia the gap ceased to narrow after 1940 and in the case of Venezuela the watershed year is 1960 (from 1950 to 1980 Venezuela was the richest country in Latin America). Ecuador and Panama narrowly closed the gap on the United States, but the narrowing is not statistically significant in the case of Ecuador and in the case of Panama is limited to the period from 1950 to 1980. Peru, virtually unchanged over the century, is also worthy of mention here as the Peruvian data show a strong improvement up to 1960 and a sharp deterioration subsequently. Elsewhere, the differential either widened or was unchanged . Furthennore, in the case of the five success stories in 1900 (Argentina, Chile, Cuba, Mexico and Uruguay), all suffered severe losses. This was almost continuous after 1910 in the cases of Argentina and Cuba, and after 1930 in the cases of Chile and Uruguay. Mexico succeeded in narrowing the gap with the United States between 1950 and 1980, but for the rest of the century suffered a widening of the differential, ending substantially below its position at the beginning.

Some of the losses are uuly tragic. Haiti , with 'a GOP per head in 1950 only 3.9 percent of the US figure , experienced an almost continuous decline and ended the period with a GOP per head of 1.1 percent of the US figure . Bolivia also suffered a continuous decline until a very modest reversal began in the I 990s. Even Costa Rica , a model democracy and one of Latin America's best­ governed countries, failed to narrow the differential on the United States. Elsewhere in Central America the ratio fell severely throughout the century and not just during the turbulent decade of the 1980s.

When sub-periods are analyzed, starting with the phase of export-led growth up to 1930, the performance of Venezuela stands out (see Table 2 .2) . The exploitation of oil under the dictatorship of Pre sident Juan Vicente G6mez helped GOP per head jump to almost 20 percent of the US figure . In Peru foreign investment and the mining boom almost doubled the ratio, while Chile - despite the dislocation of World War 1- managed to narrow the gap between them and the United Stales and Argentina. The lauer ' s perfonnance was dis­ appointing, but still gave no hint of the horrors to come.

The inward-looking development phase, based on import substitution , is con­ ventionally placed from 1930 to 1980. It brought some successes, notably in Brazil and in Mexico. Argentina and Uruguay held their own from 1930 to 1940 before experiencing a steady decline subsequently. Chile perfonned poorly and Cuba disastrously ; in both cases decline set in long before socialist economics took hold .

Economic performance since 1980 must be divided into two phases despite the short length of the period . The first is the ' lost decade' of !he debt crisis, during which all countries lost ground on the United States. Since 1990 no fewer than 12 republics have narrowed· the differential, albeit by a very small

ECOfWmic performance and lhe Slale in Lalin America 17

amount in many cases, but too much should not be read into this 7 In most cases the catching-up can be attributed to a short phase of non-sustainable economic growth based on recovery after severe recession. Examples are EI Salvador, Panama and Peru. A rew countries , however, appear to have moved towards a new and sustainable growth model capable of narrowing the differential with the Urtited States. Argentina and Chile are the most obvious cases, but Bolivia may also be in this category. Sadly, neither Brazil nor Mexico can yet be placed in this group with any confidence.

THE STATE AND ECONOMIC DEVELOPMENT IN LATIN AMERICA

There is a vast literature on the role of the state in Latin America. However, much of this literature explores those partS of the state's activities that do not impinge directly on econontic penormance. Examples are the relations between the state and civil society, the transformation of subjects into citizens and the establishment of electoral and political rights. These non-economic functions of the state are very important and will undoubtedly have indirect implications for the economy. Nevertheless, they are not central to tbe argument in this chapter and will not be pursued further.

The starting point for an analysis of the state's role in Latin America has to be an examination of the dominant economic system. With minor exceptions the dontinant system in Latin America in the last century has been capitalism. The exceptions (Cuba since 1959 and Chile from 1970 to 1973) are well known and need not detain us further. It should, however. be clear that the role of the state in a socialist system - sucb as in Cuba - is quite different from the role of tbe state in capitalism. It is not possible to develop a theory of the state for the two cases except at the most ~lllI"rflciallevel.

The state under capitalism has many functions. In order to carry out these functions, several preconditions Illust be met. These preconditions are usually taken for granted in analysis of the state in developed market economies. but they need to be specified clearly in the case of Latin America. The first pre­ condition is the consolidation of the nation state and the establishment of territorial control. This has now been largely achieved in all Latin American countries, but it was not the case at the start of the 20th century. The large Indian populations in countries such as Bolivia. Guatemala and Paraguay were treated as second-class citizens by the ruling elite and did not in many cases see themselves as part of the nation; and the territorial integration of many large republics was not assured before the advent of civil aviation and the widespreadl use of roads for travel. Civil war, a condition now found only in Colombia, has

Liberalization and its consequences 18

disrupted the nation state in many Latin American counlrie.s this century and bas undermined the ability of the state to fulfill its functions within capitalism.

Second, the state must exhibit a certain autonomy from the private sector elites that dominate the economy under capitalism. At the beginning of this century, when the state was usually under oligarchic control, there was often lillIe or no autonomy in state action. Indeed, in the smaller states the state oligarchy and the private sector elite were often the same people. Where the state is too close to private sector interests, it can generate rapid growth - as in Nicaragua under the Somoza dynasty - but it is not sustainable; the inability of the state to mediate between conflicting interests eventually leads the losers to resort to force. As the oligarchic state in Latin America has given way to the bureaucratic state, marked by high levels of public sector employment, the state's capacity to act independently of private sector interests has increas.ed. However, this is no guarantee of rational behaviour as the state may become captive to the interests of its own employees. This inertial state has been par­ ticularly problematic in three of the region's longest-standing democracies ­ Costa Rica, Uruguay and Venezuela - where the link between political parties and the state bureaucracy has been particularly damaging.

The third precondition is the establishment of the rule of the law. This is in many ways the most important precondition and yet it is the furthest from being universally achieved. The establishment of institutions capable of enforcing and upholding the legal system is a goal to which many republics in Latin America can only aspire. Far too often wrong-doing is not punished and there are no sati sfactory mechanisms in place for resolving disputes. To these traditional problems in the application of the rule of the law, a new one ­ personal security - has now been added. Fear of robbery, kidnapping or murder is a serious obstacle to rational economic decision-malcing.

The preconditions outlined above must be met if the state is to carry out its functions effectively. Of course, the state must still attempt to meet the demands placed upon it even when the preconditions are not met, but its capability to do so is likely to be ilighJy constrained. In what follows I shall therefore assume that the preconditions have been satisfied.

The first function of the state is the provision of public goods, i.e. those that are non-rival and non-exclusionary.8 The most obvious example of a public good is defence and this has indeed occupied the attention of all Latin American states. However, servicing the external public debt is another public good that has absorbed a higb proportion of public expenditure in Latin America. There are two other public goods that deserve special mention in the Latin American context. One i.s macroeconomic stability, which is correctly defined as a public good as its consumption by one citizen does not reduce the supply available for others and users cannot be prevented from consumil.lg the good. The other is measures to protect or improve the environment, although some such

Economic pelformance and the stale in Latin America 19

measures could possibly be excludable. Defence, external debt servicing, macro­ economic stability and environmental protection are not the only public goods of interest to Latin America. However, they do absorb a high proportion of government expenditure and administrative effort. Latin America's record in the provision of public goods has been mixed with only defence consistently receiving a high priority. Yet in many cases defence expenditure has been used for internal repression rather than the protection of territorial integrity, distorting its definition as a public good.

The second function of the state is the provision of semi-public goods. These are goods th.at are eitber non-rival and excludable (club goods) or rival and non­ excludable (common ,property goods). The most important are education, health, social security provision and physical infrastructure. Expenditure on semi-public goods now accounts for a very higb proportion of total public expenditure in Latin America (see Table 2.3) - as in developed countries - with serious problems in the quality of the output obtained from this level of spending.

The third functi on of the state is the provision of private goods - usually through state-owned enterprises (SOEs). This function bas now fallen out of favor , as in developed countries , but a residue remains in all republics. It was not important in the first half o f the 20th century and it was originally seen as a response to market failure , i.e. the state provision of private goods and services was seen as the appropriate respon se to some perceived failure of the pri vate sector.

The fourth function of the state is a concern with equity. The state has always been responsible for the alleviation of extreme poverty (including famine and disaster relief). However, it is also widely accepted that the distribution of income emerging from a private market system may be sub-optimal. Thus, all states use a system of transfers and subsidies to redistribute income. While this secondary di stribution of income is supposed to be more equitable than the primary distribution of income, this may not be so if either the state has been unduly influenced by strong pressure groups or if the state bureaucracy is able to thwart the intentions of governments in this area. This has been a serious problem in Latin America for most of this century.

These functions can only be satisfied with public revenue and the capture of resources to fulfill state functions is the fmal state activity to be discussed here. Ideally, the state would acquire its resources in a non-distorlionary way ; however, there are almost no taxes that do not alter relative prices and therefore affect the allocation of resources. Tbus, the impact of public taxation on the private sector needs to be taken into account in any discussion of the state under capitalism.

At the beginning of the 20th century , all Latin American states relied heavily on income from trade taxes (import tariffs and export duties). It has been estimated? that the average for Latin America was 64 percent, that is, nearly

Liberalization and ils consequences20

Table 2.3 Structure of central government expenditure', 1991-95 (0/0)

Public Goods Semi-public Goods Olber

Defence Interest paymentsb Heallb Education

Social security & welfare

Argentina 6.8 10.2 2.2 7.0 47.5 26.2 Bolivia 9.5 8.3 6.3 18.0 15.3 42.5 Bnu;il 2.8 44.5 5.9 3.5 30.6 12.8 Chile 9.2 6.3 11.4 13.5 33.5 26.0 Colombia 8.0 10.0 5.4 18.3 8 .3 50.1 Costa Rica 15.2 26.7 21.0 14.4 22.8 Dom . Rep. 4.9 6.1 11.0 10.1 3.7 64.3 Ecuador' 12.2 17.8 9.0 24.9 1.4 34.8 El Salvador 14.0 13.2 8 .0 13.4 4.5 46 .9 Guatemala 13.9 11.4 10.1 16.8 4.4 46.8 Hallie 3.4 Honduras Mexico 3.8 22.7 3.0 23.7 21.0 25.9 Nicaragua 7.6 10.7 13.4 15.4 16.3 36.6 Panama 5.2 9.7 20.5 18.4 22.4 23 .8 Paraguay 12.1 7.0 6.3 17.6 15.6 41.5 Peruc 20.1 21.6 5.7 16.9 35 .6 Uruguay 6.5 5.9 5.4 6 .7 58.4 17. 1 Vene:ruciaC 6.3 12.4 8 .6 18.3 6.7 47 .8

Nores : - = not available. • Excludes expenditure by provinces (states) and municipalities (this helps (0 explain low figure

for education in Brazil). b Includes payment on internal debl (nol a public good). c 1981-90.

Sourct: Derived from World Bank (1997).

two-!hirds of all revenue came from trade taxes. This ratio was still aroundl 40 percent in 1930 and 20 percent as lale as 1950. This dependence on trade taxes was a reflection bo!h of the importance of foreign trade and of the ease of collection. However, trade taxes are never neutral and the impact on the allocalion of resources - not 10 mention rent-seeking - has heen considerable. The distortions associated with trade taxes eventually became excessive and were one of the reasons for the abrupt shift to trade liberalisation after !he 1980s . The state supplemented taxes on trade with taxes on domestic goods and services (sales taxes). Since these were typically levied at a non-uniform rate or were subject to exceptions as well as panial evasion, they also had a major impact on !he allocation of resources . They were also nbt applied to !he

21Economic performance and the stale in Latin America

informal sector, which is one reason why the latter has grown so rapidly in many Latin American countries.

From the 1930s onwards the state carne to rely increasingly on social security contributions. Since these were collected on a pay-as-you-go basis, they should be viewed as a tax on labor in the formal sector. Demographic changes, coupled with populist legislation, have made the social security system technically bankrupt in many countries and have forced a reconsideration of pension provision in particular. Income taxes, including taxes on profit and capital gains, also became important from the 1930s onwards. Although they have never represented the same proportion of revenue as in developed countries, at their peak in 1970 they accounted for 36.5 percent of the total. Most of this was attributable to large-scale enterprises where the tax burden could be reduced through capital expenditure. Thus, income tax has also been non·neutral in its impacl on the allocation of resources.

Table 2,4 Structure of central government revenue', 1991-95 (%)

Income, Social Domestic profit and security goods and Foreign Other Non·tax

capital gains contributions services trade taxes revenue

Argentina 2,6 45 ,9 26.6 7,9 4.6 9 ,1 Bolivia 4.3 7,5 38.5 6,6 0.5 33,9 Brazil 16.3 30.l 19,2 1.9 5,2 27.2 Chile 18.1 6 .7 45.2 9,6 3,7 16.7 Colombia 39.3 37.2 9 .7 0.7 13.2 Costa Rica 9.8 27,7 32,2 16,1 0,7 13,1 Dom. Rep. 17.1 4. 1 25.1 44.2 0.3 8.6 Ecuador 56 ,0 24.9 11.6 1.0 5.5 EI Salvador 22.2 49 ,8 [7,0 0,2 7.1 Guatemala 19,4 42.4 20.1 3.5 13,7 Hailib 14.9 0.3 30,7 23.8 6,2 11.5 Hondurasb 24.2 259 42.4 1.0 5.7 Mexico 34,2 17.9 50,2 7 .1 2 ,4 9 ,0 Nicaragua 11.2 11 ,8 45.1 19,4 5.3 M Panama 18 ,0 20,5 16.8 10.5 1.7 31.1 Paraguay 9.8 27,7 14,8 11.9 30,0 Peru 12,5 12.3 493 9.8 7 .3 8,6 Uruguay 7 ,3 30,1 32,4 5.6 8.4 5.9 Venezuela 50.6 5,5 10,4 9,1 1.3 22.3

Notes: - "" not available. a Excludes provincial (state) and municipal revenues. b 1981-90.

Source: World Bank (lm). Table A.l . p. 196.

22 Liberalization and ilf consequences

The overall sbUcture of state revenue in the 1990s is shown in Table 2.4, where it can be seen that trade taxes are now of negligible importance while indirect taxes (including value added tax es) are dominant. What the data do not show is the impact of the inflation tax, which was used by the state in many countries after 1940 to transfer resources from the private to the public sector. As is well known, the 'inflation tax' is paid primarily by those who rely on cash rather than interest-bearing deposits and is therefore a tax On the poor. The reduction in inflation in Latin America has substantially reduced the importance of this inequitable and regressive tax.

ECONOMIC PERFORMANCE AND THE STATE­ THE SUPPLY-SIDE

The growth of the economy can be seen from either the supply-side or the demand-side. To analyze the supply-side. a technique known as growth accounting has been developed. This disaggregates growth in any period into a contribution from the growth of factor inputs (land, labor and capital) and the growth of total factor productivity (TFP), that is the growth in output per unit of factor input with all factors of production aggregated together. A few attempts have been made to apply growth accounting to Latin America lO The results are not very reliable in view of the quality of the data, although they do suggest the limited importance of 1FP growth and the greater importance of the increase in factor inputs. What can be done, however, is to apply the growth accounting approach to economic performance in Latin America to see what light it sheds on the role of the state. In order to do this, it is necessary to resolve how to treat the change in the quality - and not just the quantity - of factor inputs. In the simplest versions of growth accounting, this qualitative cbange is treated as part of TFP. However, it is al so possible to treat the change in quality as a separate variable in the explanation of growth. Thus, for the three factor inputs we have to consider both the change in quantity and the cbange in quality , while continuing to include 1FP as the residual in the growth accounting equation.

By the beginning of the 20th century, the state had already distributed a large proportion of the public land it had inherited from the Spanish or Portuguese crown. This had vastly increased the land in private ownership and led to an accumulation of productive land inputs. This process continued after 1900, with many governments offering land on very favorable terms to new immigrants. Deforestation played only a small role in the growth of land devoted to agriculture in the frrst half of the cenrury. However, it became increasingly important after 1950 as demographic pressure combined with the tax system to provide in centives for the encroachment of agricultural land on the primary

Economic performance and the stale in Latin America 23

and secondary forests. By the 1980s it had become apparent that the tax system was producing major distonions in the choice of land use, panicuJarly between forests and pasture, but deforestation has not yet ceased because the private rate of return to non-forest use remains higher.

The final way in which the state has affected the accumulation of land inputs is through land reform. Although much of this involves a redistribution of land, the theory behind it assumes that a larger proportioo of land held in large units is idle than in small units. Thus , land reform should lead to an increase in land inputs. However, land reform in Latin America has been a big dis.appointment for those who hoped that it would improve equity. However, it did accelerate the transition to modem capitalist agriculture, leading to a big increase in agri­ cultural output and labor productivity.l' Indeed, agricultural labor productivity growth (see Table 2.5) seems to have peaked during the years in which land reform was being implemented.

Table 2.5 Agricultural labor productivity: average annual growth role (%)

Argentina Brazil Chile Colombia Mexico Venezuela

1st Period 1.7 0.8 0.6 1.3 (1903-52) (1921-48) (1909-26) (1901-10)

2nd Period 2.6 0.8 2.1 0.9 1.2 (1947-70) (1927-41) (1926--66) (1922-47) (1926-50)

3rd Period 3.9 5.9 3.3 3.3 6.0 (1953-79) (1971-81) (1942-71) (1944-70) (1951-80)

4th Period -<).7 1.1 1.8 3.0 0.6 -1.1 (1980-90) (1982-90) (1973-90) (1967-93) (1971-90) (1981-90)

Overall 2.1 2.2 1.9 2.4 1.6 3.1

Source: Thorp (1998). Appendix Table JV.l. pp. 32&-9.

The state bas therefore played a key role in the growth in the quantity of land inputs. What about its contribution ID the quality of land inputs? Here the picture is more mixed. State-sponsored irrigation schemes have been the most imponant source of an increase in quality, but these have oot been very widespread. t2 Meanwhile, the tax system has encouraged the spread of agricuilure to lands that are oot well-equipped for this purpose - the Brazilian Amazon being the main, but not the only, example.

The fiscal system bas also had a strong urban bias for much of the 20th century as a result of the policy of import-substituting industrialization. The agricultural sector often received zero tariff protection, but paid high tariffs on its inputs. At the same time, price control on foodstuffs lowered its relative

Liberalization and jts consequences 24

price and encouraged a shift of resources out of rural areas. The worst distortions in the tax system had been corrected by the 1980s and trade liberalization has now led to a tax system which is more neutral between agriculture and other activities. With a few exceptions, such as EI Salvador and Haiti, there is no evidence that the growth of aggregate output was constrained by a shortage of land. Thus, tbe state must be judged to have fulfilled its functions in this area. The same is not true of labor inputs, where shortages were endemic until the fall in the crude death rate after 1930 rurned Latin America's labor shortage into a labor surplus. The state's preferred response to labor sbortages was legislation to encourage immigration, preferably from Europe. Only a handful of republics - Argentina, Brazil, Chile, Cuba and Uruguay - had mucb success in attracting inunigrants and only in Argentina, Cuba and Uruguay did it amount to a significant share of the labor force. Where voluntary migration from Europe failed, the state resorted to indentured labor from Asia or coercion of domestic labor. The lauer was achieved primarily by denying part of the rural population access to land, thereby forcing them to work for wages. These short-sighted policies help to explain the stagnation of real wages in Latin America at a time wben labor productivity was rising. The stagnation of real wages in turn undermined the growth of the internal markel, making it difficult for the export sector to stimulate the growth of the non-export sector.

From the 1940s onwards there has been no labor shortage in Latin America. Thus, the growth of output in the second half of the century has not been constrained by labor supply. The quality of labor inputs, however, is anoLber story. Studies of contemporary Latin America have emphasized the educational deficit among Latin American workers compared with other developing regions.13 Since mass educational spending was not until recently a priority in most Latin American countries (the main exception is Costa Rica - see Table 2.3), it is not unreasonable to assume that the educational deficit existed in earlier years. Much the same is true with regard to healtb spending, which has an important bearing on the quality ofiabor inputs.

If the state can be accused of negligence with respect to the quality and quantity of Jabor inputs, the same is not true of capital inputs. On the contrary, it can be argued that the state has at times encouraged excessive accumulation of capital at the expense of employment and with insufficient auention to the social rate of return on investment. At the beginning of the century, the main instrument used by the state to promote capital inputs was legislation in favor of foreign investment. This included generous - often excessively generous­ concessions to foreign companies through guaranteed minimum rates of return Of tax holidays. These policies were supplemented after the Great Depression in two ways. First, the tariff system favored the import of capital goods where no capital goods industry ex.isted; second, development banks were established to channellong-tenn finance to projects that failed to find favor with private

Economic perjormonce and the slate in Latin America 25

commercial banks. The result was a substantial increase in the accumulation of capital inputs - not all of which was economically rational.

The state's role in promoting the growlh of capital inputs was further advanced with Ihe explosive growth of SOEs after 1950. Originally justified by reference to market failure, SOEs were located above all in the capital-intensive branches of the economy where they were responsible for up to half of all gross fixed capital formation. In the larger countries (Argentina, Brazil and Mexico) they were also mainly responsible for the growth in the public external debt after the ftrSt oil crisis in 1973.14 The state has therefore played a major part in promoting the accumulation of capital inputs. However, the quality of this increase has left much to be desired. The tax system gave a capital-intensive bias to much private sector invesunent, while the financing of capital inputs in SOEs became a vehicle for capital flight in the 1970s and 1980s.15 The quality of capital inputs in Latin America bas been disappointing fOf.much of this century.

The growtb of TFP in Latin America, particularly when stripped of the change in th.e quality of factor inputs, has almost certainly been modest, although empirical data are not strong enough to confirm this. TFP, until recently, was treated as exogenous, although the new growth Iheory now views it - probably correctly - as largely endogenous. t6 TFP is influenced by many things. First is the legal framework to encourage innovation, where it is widely assumed that a strong patent law is important. The legal system in much of Latin America has been deficient with respect to patent protection in general and intellectual property in particular, although this is now changing under pressure from the World Trade Organization. Innovation has been far less important than imitation in Latin America, but- unlike in post-war Japan ­ imitation did not lead on to innovation . One reaSOn is the absence of competitive markets. Shielded from impons by high nominal tariffs and unable to export because of anti-export bias, industrial firms in Latin America have operated in oligopolistic markets where innovatioD was not necessary to sustain prof­ itability. Even Brazil, with its vast intemal marke~ suffered from this problem, although the Brazilian state did at least benefit from research and development expenditure by foreign companies that required some response from local flnns.

The third stimulus to TFP growth is economies of scale. As Ihe market expands, the unit cost of production falls and total factor productivity rises . Economies of scale are primarily found in the manufacturing sector, which is related in tum to the size of Ihe internal market under import substituting i.ndu.s­ trialization (lSI). It is nO accident, therefore, that Brazil has been the biggest beneficiary of TFP growth induced by economies of scale, although it is by no means clear that this has anything to do wilh the role of the state.

The state in Latin America has therefore been successful in preventing land from constraining growth , but less so in the case of labor during the period of labor shortages . The state has contributed to the rapid growth of capital inputs,

26 LiberaUzation and ils consequences

particularly after World War II, but the quality of these inputs has left much to be desired. Finally, the state bas played little or no role in the growth of TFP; this is in marked contrast to developed countries throughout this century or a number of countries of Southeast Asia since the 1950s. The most important exception has been the state in Braz.il. Although the growth of capital inputs ­ at least in the formal sector - may have been excessive, the Brazilian state has played a key role in facilitating the growth of all factor inputs throughout this century and has even played a modest role in promoting the growth ofTFP. In this respect it has benefitted from the size of the internal market - a key determinant of growth in an inward-looking model of deveiopmellt Although the tax system has created many distortions in Brazil, the autonomy of the state allowed bureaucrats to design policies that partially compensated many of the losers. This was far from ideal, but it did help Brazil to close the gap On the Urtited States.

ECONOMIC PERFORMANCE AND THE STATE­ THE DEMAND-SIDE

Growth accounting can be applied to the demand-side as well as the supply-side. Using techrtiques pioneered by Hollis Chenery in the 1960s, it is possible to calculate the contribution to growth in real GDP of import substitution on the one hand and on the other the components of final demand (private consumption and investment, public consumption and investment, and exports) on the assumption of no change in the import coefficient. 17 Let us start with import substitution. II is often assumed that import substitution refers only to manu­ facturing or perhaps industry (that is, including mining), but it cao include any tradeable sector where importS are significaot. Thus, in some countries in Latin America import substituting agriculture (ISA) has been important and in the region today there is negative import substitution in services (ISS) as a result of the replacement of previously non-traded activities (for example, financial services) by imports.

With this proviso the main focus this century in Latin America has been import substituting industrialization (lSI). However, mucb lSI is 'natural' and lakes place without state intervention. As the market widens, finns with increasing returns to scale (common in manufacturing) are able to Ilower costs of production and replace part of the import supply. This natural process was well under way in the larger Latin American republics before the 1930s. The state, however, plays a key role in import substitution by selecting the rate for tariffs and many non-tariff barriers (for example, quotas). Raising these barriers creates new' artificial' opportunities for import substitution even if the state

Economic perjomUJnce and the stale in Latin America 27

increased tariffs and quotas for other (that is, revenue) reasons. It was only in the 1950s that most Latin American states embraced a tariff and noo-tariff barrier (NTB) strucrure that was overtly protectionist.

As a device for accelerating industrialization, state-sponsored lSI was highly successful. The share of industry (or more precisely manufacturing) in Latin America rose sharply from the 1930s to the debt crisis in the early 1980s and the contribution of import substitution to the growth of manufacturing was almost 50 percent. However, the protection offered was excessive and generated several distortions. First, it turned the internal terms of trade against the agri­ cultural sector and, second, it created a strong anti-export bias making it difficult for even successful firms to sell their production outside the national market Finally, by limiting imports it reduced competition in the home market and contributed to the problem of chronic inflation in the larger countries. The state­ sponsored regional integration schemes adopted in the 1960s were too timid to make much of a difference, and the problems were only addressed with the adoption of trade liberalization and bolder integration schemes in the wake of the 1980s debt crisis.

Private consumption in Latin America is the largest component of final expenditure and absorbs a very high proportion of private incomes. It is a function not only of the level of income, but also its distribution. Thus, the high degree of income inequality in all Latin American republics is, and has been, reflected in the pattern of consumption. The state has rarely tried to fine-rune private consumption - a very difficult task in view of the modest impact of interest rates on household spending until recently in Latin America. However, the state's reliance on net indirect taxes (that is , net of subsidies) has substan­ tially affected household expenditure. Subsidized items, such as tortillas in Mexico, usually form a substantial part of the expenditure of low-income families, while the impact of tax rates depends on the price elasticity of demand;. where it is rugh, contraband is often the consequence and this is one of the reasons why governments in the region have adopted more modest rates of indirect tax (including tariffs).

The income of households is either consumed or saved. Thus, the level of private consumption also depends on the incentives for private savers. These have been very unsatisfactory for much of this century as a result of a lack of financial instruments for private savers and frequently negative real rates of interest. Thus, despite the unequal distribution of income, Latin America has had low savings rates - the worst of both worlds. IS The most important impact of the state on private consumption has probably been through the tariff on importS of consumer goods. Before 1930 a high proportion of importS consisted of consumer goods; by 1980 this was close to zero in the larger republics so that almost all private consumption was met through domestic supply. With trade liberalization in the 1980s and regional integration in the I 99Os, the proportion

,

28 Libt ralization and its consequences

of imports consisting of consumer goods has risen sharply; this negative import substitution has lowered celeris paribus the contribution of private consumption to GDP growth. Recently, with the increase in the importance of durable goods consumption, credit cards and personal credit, the state has begun to refme its approach to the management of private consumption. This has been most apparent in Brazil since the launch of the Real Plan in 1994, where measures have been laken at regular intervals to curb private consumption. Brazil is not alone in this, however, and many Latin American states are now closer to the European or North American model in whicb household consumption is sensitive to the rate of interest as well as to the level of disposable income.

The role of the state in the promotion of private and public investment has already been discussed (see previous section); and the accumulation of capital inputs on the supply-side has its mirror image in investment expenditure on the demand-side. It is worth stressing, however, the role of SOBs in sustaining high rates of investment during the final two decades of the import SUbstitution pbase. The private sector almost never achieved a rate of inveslrnent in excess of 20 percent of GDP and in some cases it was less tban 10 percent. State companies, with their privileged access to the domestic and international capital markets, were responsible for between one-third and one-half of gross fued capital formation in the larger republics before the debt crisis. Today, however, public investment, both by government and by SOEs, is now much lower as a proportion of GDP. The private sector is being asked' to cany the main burden of investment, but domestic real interest rates remain high. Thus, access to foreign capital markets through bond issues, bank borrowing or international equity issues has become much more importanL The state has an important role to play here since the terms on which private ftrms access the international market are affected by the foreign perception of country risk. A low rating imposes additional costs on all investments, not just those carried out by state agencies. In terms of growth accounting, what maners is the domestic component of investment expenditure, that is, after excluding imported capital goods. The capital goods share of imports is high in all republics and in most countries - Brazil is the main exception - the domestic component of investment in machinery and equipment is low. This, however, is nothing new and is a reasonably accurate description of Latin America throughout this century.

Public consumption is an important component of final expenditure and its import content is low because it is dominated by wages and salaries. It has grown faster than GDP this century, as in almost all countries in the world, both because the demand for public services increases with urbanization and the rise in GDP and because the state now has duties (for example the mailllenance of an airforce) that did not exist at the beginning of the century. Indeed, it is probable that public consumption's share of GDP has doubled or even trebled in the last 100 years. Unfortunately, it is not possible to compare public consumption over time

Economic performance and the stale in Latin America 29

for all Latin American countries. We do know, however, that in general the proportion devoted to health, education and social security has risen and that the proportion spent on defense and debt servicing has fallen. The fall in debt service costs may seem surprising, until one remembers the small size of total public consumption at the start of the century in absolute tertns.

Growth accounting can show the direct impact of public consumption on GOP growlJl, but how effective has the growth of public consumption been in the promotion of GOP growth? The indirect effects come in several fortn_s ­ positive and negative. The positive impact comes from the increase in private sector factor productivity made possible by greater public expenditure. Before World War II, it was widely assumed Ihat spending on infrastructure had the greatest impact; in more recent years, with the private sector playing a greater part in the provision of physical infrastructure, attention has shifled to health and education. There is no doubl that, with a few exceptions, slate spending on health and education has not brought the expected benefits for much of the post-war period. There have also been indirect costs. In developed countries the problem has often been seen as one of 'crowding out' private spending through the impact of public borrowing on interest rates. In Latin America this has been much less serious because the economies are less sensitive to interest rates. However, the state has relied on indirect rather than direct laxes to pay for public spending and this has often distorted relative prices and incentives. This has been the most serious negative effect of increased public spending in Latin America and examples have been given above. Furthertnore, the growth of public spending has also led to a huge increase in the size of the state bureaucracy. It has been estimated t9 that public employment as a proportion of total employment rose sharply between 1950 and 1986 al a time when total employment was itself rising rapidly. In Costa Rica, for example, it rose from 2.5 to 19.8 percent, in Uruguay from about 9 to 20.7 percent and in Venezuela from 6.7 to 19.3 percent. These public sector bureaucracies have become an interest group in their own right, often thwarting the implementation of government policies and resisting administrative reform.

The final contribution to GOP growth on the demand-side comes from exports. At the beginning of this century public policies laid great stress on the promotion of exports. Those countri,? that out-performed the United States before 1930 (see Table 2.2) did 'so in most cases as a result of a fast growth of exports. Foreign direct investment played a major part in this export expansion (for example, Venezuela after 1917), but it would be wrong to underestimate the part played by the state through the legal framework, infrastructure spending, the tax and subsidy system as well as other incentives. In Colombia, for example, the outperformance of the United States before 1930 was clearly linked to the rapid growth of exports of coffee and bananas where, in different ways, the state's role was very important. After the 1929 depression the state

30 LiberaLization and ilS consequences

struggle<! to suppon the expon sector in the face of the decline in world trade and the collapse of commodity prices. Public policies were not unsuccessful and the volume of expons (if not the value) recovered quickly after 1932 in most countries. However, the mOVe to a fully-fledge<! model of inward-looking development after 1950 cripple<! the export sector in most countries. As the anti-expon bias inherent in the structure of protection began to take effect, Latin America saw its share of world trade decline sharply.

The new growth model , base<! on market-friendly policies and low trade barriers, has create<! new opportunities for the expon sector.20 Latin America, however, has not retume<! to expon-led growth. With the exception of Mexico, where NAFT A provides unrivalled opportunities for export expansion, the new growth model has eliminated the worst distortions faced by the expon sector, but has not produced' a consistent policy of expon promotion. Policy is now more neutral between exponables, imponables and the non-traded sectors with the allocation of resources determined primarily by relative prices set in free markets. This is a far cry from state-directed development, as occurred in many pans of Southeast Asia, but it is a big improvement on the anti-export bias found before the debt crisis and may be the best that can be achieved with Latin America's badly trained and poorly motivated public sector bureaucracies.

ECONOMIC PERFORMANCE AND THE ROLE OF THE STATE: CONCLUSIONS

The 20th century in Latin America divides relatively easily into three periods. In the first, 'up to the end of the 1 920s, Latin America continued to pursue the model of export-led growth developed in the second half of the previous century. In the next period, up to the first oil crisis in 1973, Latin America followe<! an inward-looking model of development with a heavy emphasis on industriamation . In the fmal period, which began in the southern cone after 1973 and in the rest of the region after the 1982 debt crisis, Latin America adopted a neo-liberal model in response to globalization, e)[ternal shocks and the perceived exhaustion of the inward-looking model.

The role of the state in Latin America has evolved as the economic paradigm has changed. The secular increase in resources at the disposal of the state has belpe<! to resolve some problems, but at times the SUite has been unable to respond adequately to the new challenges it bas faced. The state has also become more effective in meeting some of its functions and less effective at addressing others.

In the first period, marke<! almost everywhere by an oligarchic state commanding few resources. the preconditions for an effective state were often

Economic perfonnance and the state in Latin America 31

absent. Territorial integrity was not assured in many countries, the state lacked autonomy from the private sector elites and the rule of law left much to be desired. The main exceptions (Argentina, Chile and Uruguay) were all in the southern COne. In Mexico the state was too closely identified with private sector interests and in Brazil the control of the federal government over the provinces was minimal, leaving the authorities unable to implement an effective state policy. In many of the smaller states of the Caribbean basin, the 'state ' consisted of representatives of the US government, as dollar diplomacy was applied to resolve the vacuum created by weak domestic rule. During this period the priority for the state was the promotion of the export sector, where oligarchic interests were concentrated. These efforts were highly successful in the period until World War I as the state cooperated closely with the private sector. However, the war served as a reminder of the vulnerability of Latin American economies to external shock and this fact was driven home further by the .collapse of commodity prices in the 1920-21 global depression.

While the expon sector boomed, macroeconomic stability was assured and the public debt was serviced without difficulty. Before the war this was underpinned in many countries by the gold standard, the silver standard or even the dollar standard. However, monetary arrangements were disrupted by the war and macroeconomic stability suffered accordingly. The state in this period devoted a high proportion of its scarce resources to defense and a modest aroountto semi-public goods such as health and education. Only Uruguay, the pioneer of the welfare state, devoted resources to social security and the provision of private goods through SOEs was minimal. Neither protection of the environment nor promotion of equity were taken into account at tlus time. Even if they had been, the tiny share of GDP captured by the state would have ruled out any major initiatives. The state in tbis period was effective in promoting export-led growth, but suffered from two main weaknesses. First, it had no response to external shock, which could easily overwhelm its best efforts to encourage international trade; only Brazil, with the adoption of a valoriza­ tion scheme for coffee, made a serious effort to address this problem and even in this case the initiative did not initially involve the federal government. Second, the close ties between business and government prevented the state from making a serious attempt 10 diversify the economy away from excessive dependence on a small number of export products. The region paid a heavy price for such dependency during the Great Depression.

In the second period, the paradigm shifted 10 an inward-looking model of development. The state, with a few tragic exceptions such as the Chaco War between Bolivia and Paraguay in the 1930s, was no longer troubled by an absence of territorial integrity . The federal government in Brazil, particularly after Geullio Vargas instituted his Estado Novo (New State) dictatorship in 1937, gained greater control over public policy; meanwhile, President

32 Liberalization and its conseqlU!nces

Roosevelt's Good Neighbor Policy encouraged dictators in the Caribbean basin to fill the gap left by the withdrawal of US troops. All these states began to distance themselves from the traditional eliles that had dominated decision­ making before Ibe 1930s and Ihis process was carried further with the enfranchisement of a greater proportion of the population.

Autonomy did not necessarily bring accountability, but the growth of state resources strenglbened tbe rule of law. Banditry declined, the judiciary was strengthened and the state acquired the ability to negoliale on more equal terms with foreign inveslors. These encouraging trends reached a peak in the 1950s, after wbicb the rise of guerrilla movements , the spread of mililary regimes and a growing hostility 10 foreign companies once again raised doubts about the rule of law in many countries. The preconditions for an effective stale were therefore closer 10 being mel during the second period than ever before. In addition, the prevailing global ideology - heavily influenced by experience in World Was II - favored a strong state and the new international agencies were able to support many government activities. The most importanl was investment in infrastructure, where the state set aboul cOFlstructing the transport, energy and finance systems needed to support industrialization.

The shift to an inward-looking model of development, with its inevitable anti-export bias, placed a strain on the balance of payments and led to chronic inflation in Ibe lasger countries. Smaller countries, where the export sector was not so neglected , suffered much less. AlmoSI all countries wasled resources on defense and not a single state put in place a system of protection for the environment despite the mounting evidence of rural degradatioD and urban pollUtiOD. The growth in reveDue allowed the state to address the shortcomings in public bealth and educatioD. Spending on social security also rose, although the beneficiaries were usually restricted to public sector employees and a small number of private sector workers. However, it was in the area of private goods provisioD thrOUgil SOEs that slate intervention differed most clearly from the period before 1930.

The case of Brazil can be used by way of illustration. From President Vargas onwasds, the state in Brazil was distinguished by its commitment to develop­ mentalism . While many things changed with military rule ill 1964, this commitmenl remained the same. The stale used public investment to promote those sectors Ibat were considered crucial to development, pasUcularly transport equipmeD~ and an alliance was forged wilh labor and capital (both domestic and foreign) to secure public goals. The fITSt oil crisis in 1973 exposed the vulnerable side of developmental ism in Brazil and the state was unable to jind a suitable response. Elsewhere in Latin America disillusionment with state-led development was already apparent by the end of the 196Os. All countries looked for an exit from the constraints of inwasd-looking development: some promoted non-traditional exports, others emphasized structural reforms and almost all

Economic performance and the state in Latin America 33

experimented with regional integration. None of these policies achieved much success and most policymakers were at a loss to respond to the rapid growth of the East Asian tigers.

The question has often been asked if Latin American countries could have replicated the role of the state in East Asia during this phase of its development. The answer is almost certainly in the negative. The East Asian state (for example, in South Korea and Taiwan) was very authoritarian and the presence of an external aggressor gave the state a legitimacy that would have been difficult to repeat in Latin America. In addition, the civil service in the East Asian countries had a different tradition and ethos, being much less politicized and more subject to executive authority. Finally, the state in East Asia was seeking to promote exports while in Latin America lSI was the goal. Thus, the state in the two regions did not even share the same objectives2t

From the 1950s onwards, the issue of equity came onto the agenda for the first time. Survey data confirmed what had been long suspected: Latin America had the most unequal distribution of income and wealth of any region in the world. A few countries, notably Colombia and Costa Rica, made a serious effort to tackle inequality through the tax and subsidy system, but most governments took refuge behind the Kuznets curve, which purported to show that inequality will at first worsen with development before eventually improving.

The rise of the developmentalist state during this second period was not without its successes. Between 1930 and 1970, as Table 2.2 shows, Brazil, Mexico and Venezuela were able to reduce the gap in living standards with the United States. Yet it is surely significant that no other country for which we have data in these years was able to do so and in some cases the relative decline was very severe. Honduras and Cuba (first under capitalism and then under socialism) suffered the most, but the southern cooe countries were also very badly hit.

The final period , since 1973, marks the long transition over a quarter of a century to neo-liberalism. While there are similarities between this period and the one at the beginrting of the century, it would be an error to identify the new paradigm as export-led growth. The reason is that the state now sees its role as one of liberalizing markets so that domestic prices can move closer into line with international prices. Resources are then allocated according to these free market prices; the outcome may favor the export sector, but this is not necessarily so. The state is now supposed to be indifferent between different outcomes provided that tbey do oat threaten macroeconontic stability.

This period has been marked by the increase in the mobility of domestic capital and the rise of globalization. The private sector, in the striking phrase of Albert Hirschman, can now threaten the state with 'exit' and has done so on numerous occasions since 1973 through capital flight. 22 At the same time glob­ alization has put new constraints on the state's freedom of maneuver. The result

34 Liberalization and its consequences

is a decline in the autonomy of the state, which must increasingly take into account the interests of domestic and foreign capital. This decline in autonomy has gone hand in hand with a serious deterioration in the rule of law. 1be reasons are complex even if the result is not in doubt. Recession in the 1980s led to a rise in the underclass, which has been swollen by private sector reslrUcturing in the 1990s. Far from absorbing a portion of these displaced workers, the state has been forced to add to their numbers through administrative and SOE refonn (including privatization). The result has been an increase in violence and insecurity to which the judicial system has been unable to respond. The pre­ conditions for effective state action have therefore largely been absent in this final period. As a result, there are only modest successes attributable to the public sector. The most significant so far has been macroeconomic stability with a sharp decline in inflation rates in almost all countries. The control of inflation has benefitted from the greater autonomy given to the monetary authorities in each country and the price competition forced on domestic firms by the rise in imports following trade liberalization .

In the provision of other public goods, the role of the state has been very patchy. !)efense spending has been cut and the armed forces of many countries are moving towards a new and more relevant concept of national security. The protection of the environment has been pushed up the agenda at the rhetorical level, but state action falls far short of the required response. The external public debt, following restructuring under the Brady Plan , has been serviced in full, but Latin American states still face a hefty risk premium as a result of the failure of governments to establish credibility. The problem of credibility was much less serious in previous periods, but has become a major burden for the modem Latin American state. Credibility should therefore be added to the list of public goods to be provided by an efficient state. To achieve it, a Latin American state must not only carry out coherent and consistent economic policies, but mu st also differentiate itself in the eyes of (mainly foreign) investors from other countries both inside and outside the region. Even Chile, the most effective of the states in the neoliberal period, has found this to be a difficult task. The reduction of defense spending and subsidies to SOEs, coupled with a greater role for the private sector in the provision of infrastructure, has released resources that the state has been able to devote to semi-public goods - partic­ ularly health and education. The state, however, still faces a grave problem; first, the backlog in social spending is enormous and cannot be completed quickly; second. the state is entering new territory in which there is much learning by doing. The optimal balance between the private and public sectors in the provision of semi -public goods has yet to be found. This helps to explain the disappointing impact of neo-liberalism on equity. While poverty has in general fallen in those countries where real GDP per head is growing, the dis­ tribution of income has failed to improve. Education is widely seen as the key

Economic performance and the Slale in Latin America 35

to greater equity through its impact on the disUibution of human capital and it will take nIany years for the new emphasis by the state to bear fruit.

In conclusion, the state's conUibution to Latin America's economic performance has been disappointing in the 20th century. The state must share much of the blame for tlle failure of the region to close the gap on the United States and for the failure to improve income distribution and management of environmental resources. Yet {he new conlext gives some grounds for optimism - not so much because of the choice of a new economic Paradigm (neo­ liberalism), but because of the deepening of the democratic process. The empowerment of individual citizens, the emergence of a free press and the spread of non-governmental organizations is making new demands on the Latin American state. In the short-run, the inability of traditional political parties to respond to these demands has made the problem of state failure appear grealer, but in the long-term democratization is the best possible assurance of an improvement in state performance.

NOTES

1. Uruguay 's GDP per head in 1940 is estimated at $662 (1970 PPP prices) compared wirh$645 for Argentina. See Thorp (1998), AppendiJ: Table IX.I, p . 353.

2. The Porfuialo is the name given to lhe period in Me,uco 's hisoory (1876-1910) dominated by Porfirio Dfat. lt was a period' pf substantial economic achievement. but soc ial and poliCicaJ development lagged far behind.

3. See Bulmer·Thomas (1995), Chapter 3. 4. Cuba ceased to be a OIban colony at the end of 1898, but it was slill under US occupaCion in

1900 folloWing US ifllervem.ion in 1898. 5. J here follow the conventional definition of Latin America that excludes Puerto Rico. The

latter's level of development was similar to Cuba in 1900 and irs peformance would allow it to be classified as a moderaLe success.

6. PPP ex.change rates take into accOUnl lhe diJferen ce in the price of non·traded goods and services. These are generally much lower in Latin America (compared with the United SlaCeS) than is implied by lhe use of official exchange rates.

7. The Asian financial crisis, which began in 1997, the fall in commodity prices in 1998 and the Brazilian devaluation in 1999 have combined to lower the growth rate of GDP per head below mat in the United States since 1996.

8. See Wodd Bank ( 1997), p. 26. 9. See Tho'Jl (t998), Appendix Table V1I.I, p. 346.

10. One example is Elias (1992). The main problem in all these studies is measurement of the capital siock; this has complicated use of lhe methodology ill all regions, but in Latin America the problem is exacerbated by high I3teS o f inflation in many of the larger countries during the years to which growth accounting is applied. .

1 L See De lanvry (1990). For a good survey on land reform in the region, see Thiesenhusen (1989).

12. TIle main proble m has been the high cost and the diffi culty of ensuring that all beneficiaries contribute to the costs.

13 . 1be Inter·American Development Bank (lDB) has carried out numerous srudies in this area. See, for example. lOB (1996), ChapLer 4 .

14. See Bulmer· Thomas ({995 ), Chapter LO.

36 LiberaliUloon and itJ consequences

IS. There is a clear correlation between lending to SOEs and capital flight, although the mechanics under wltich this occurred is shrouded in mystery .

16. On the new growth theory, see Barro (1998). 17. On growth accounting applied to the demand-side, see Chenery (1960). J8. On savings in Latin America, see Grosse (1998). 19. See Whilehearl (1994), p. 398 . 20. On trade liberalization in the new growth model, see lbe chapter by FitzGerald in Bulmer­

Thomas (1996). 21. The World Bank has devoted one of its more conlroversial pUbtications 10 an analysis of the

Ea$l Asian Slate, which conlains many imeresting comparisons with Latin America. See World Bank (1993).

22. On the concept of 'exit' , see Hirschman (1981).

REFERENCES

Barre, R. (1997), Determinants o/Ecorwmic Growth. a Cross-Country Empirical Study, Cambridge, MA: MIT Press.

Bulmer-Thomas, V. (1995), The. Economic History of Latin America since Indepen­ dence . Cambridge: Cambridge University Press.

Bulmer-Thomas, V. (ed.) (1996), The New Economic Model in Latin America and its !mpact on Income Dis"ibwion and Poverty. Basmgstoke : Macmillan; New York: Sl Martin's Press.

Chenery, H. (1960), ' Patterns of indusLrial growth ', American Economic Review, 50 (3).

Colclough, C. and J. Manor (eds) (1991), States or Markets ' Neo-liberalism and the Development Policy Debate , Oxford: Clarendon Press.

De Janvry, A. (1990), The Agrarian QuestiOn and Reformism in Lacin America , Baltimore: Jolms Hopkins University Press .

Elias, V. (1992), Sources ofGrawth: A Study ofSeven Latin American Economies, San Francisco: International Center for Economic Growth .

Emmerij, L. (ed.) (1997), Economic and Social Developmenl in th e XXI Century, Washington, DC: Inter-American Development Bank.

Evans, p" D. Rueschemeyer and T. Skocpol (eds) (1985), Bringing the State Back In, Cambridge: Cambridge University Press.

Geddes, B. (1994), Politician 's Dilemma: Building S,ate Capacity in Lalin America, Berkeley: University of California Press.

Grosse, R. (ed.) (1998), Generating Savings for Latin American Development, Miami: University of Miami , North-South Center Press.

Hirschman, A. (1981), Essays in Trespassing:Economics to Politics and Beyond . Cambridge: Cambridge University Press.

Inler-American Development Bank (19%), Economic and Social Progress in Lalin America. Wasbington, DC : Inter-American Development Bank.

Klitgaard. R. (1991), Adjusting 10 Reality: Beyond 'Stale versus Markel' in Economic Development, San Francisco: ICS Press .

Maxfield, S. and B.R. Scbneider (eds) (1997), Business and the State in Developing Countries, Ithaca: Cornell University Press.

TItiesenhusen, W . (1989), Searching for Agrarian Reform in Latin America, Boston: Unwin Hyman .

Thorp, R. (1998), Progress , Poverty and Exclusion : An Economic HislOry of Latin America in the 20th Century, Washington, OC: Inter-American Development Bank.

Economic performance and lhe .state in Lalin America 37

Whitehead, L. (1994), ' Stateorganiz.ation in Latin America since 1930' , in L. Bethell (ed.), The Cambridge History of Latin Am",ica, Vol. 6, Pan 2, Cambridge: Cambridge University Press.

World Bank (1993), The East Asian Miracle : Economic Growth and Public Policy, Washington, DC : World Bank.

World B ank (1995), Bureaucra ts in Busine.ss: The Economics and Po /ili cs of Governmtnt Owner.siup . Washington, DC : World Bank.

World Bank (1997), World Development Report , Washington, DC : World Bank.