Bulmer-Thomas V., Coatsworth J., Cortes-Conde R. The Cambridge Economic History of Latin America: Volume 2, The Long Twentieth Century

Подождите немного. Документ загружается.

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

172 Alan Dye

economic decisions, at the time of the original European occupation of a

region, which produced persistent regional differences in the use of coer-

cive labor institutions. This idea emerged from an observation in com-

parative Atlantic history: though New World colonies were founded and

governed by people of European origin, until into the nineteenth century,

the Amerindian and African inhabitants were more populous. Engerman

and Sokoloff propose that geographical variation in the use of coercive labor

institutions explains the observed differences in the long-run paths of insti-

tutional development. The colonies that were most successful at extracting

labor with coercive institutions, though initially the most prosperous, were,

after centuries of institutional development, the least prosperous. Mean-

while, the growth trajectories of colonies that were least successful with

coercive labor institutions – in British North America – diverged upward

in the long run.

The key social force, in their analysis, is the inequality of wealth that coer-

cive labor institutions produced. In coercive colonies where initial wealth

disparities were large, a small wealthy elite wielded disproportionate control

of political decisions. They colluded politically to write laws, occupy gov-

ernment ofﬁces, use the coercive powers of the state to their own beneﬁt, and

exclude access by the lower classes and ethnic minorities to resources and

wealth. Even after the most brutal coercive practices were abolished, elites

in control of governments continued to erect institutional barriers to social

mobility and wealth accumulation. By contrast, in colonies where economic

opportunity and access to wealth were more evenly distributed, the early

colonizing authorities, in the interests of revenue collection, established

rule of law, made property in land and credit widely accessible, and created

formal institutions that enforced contracts with fewer class-related barri-

ers. Markets developed more autonomously and stimulated entrepreneurial

activity in a wider range of social classes, who invested in physical, human,

and social capital. Where laborers were free to pursue independent oppor-

tunities and had access to land, the dispersion of wealth was greater, and

political and economic rights were extended more broadly. Freedoms gave

common people economic and social mobility and ampliﬁed the rates of

participation and the network density of local markets.

The conditions that determined the initial use of institutions of coercion

in the Engerman-Sokoloff model were factor endowments at the time of

colonization, which determined locally optimal crop or mining specializa-

tions, labor demands and supplies, and the institutions to exploit them.

Although this component of the model has been criticized as oversimpliﬁed,

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

The Institutional Framework 173

a comprehensive explanation of the origins of the initial institutional con-

ditions would not overturn the model’s principal implication: it explains

how the basic institutional frameworks in divergent New World colonial

settings could persist indeﬁnitely without assuming they were ﬁxed exoge-

nously. Although the institutions have not remained static, their distri-

butional outcomes emerged from different formative conditions and were

self-reinforcing through endogenous decision making in path-dependent

political and legal institutions.

The main strength of their analysis is its effective challenge to old, incon-

sistent habits of thinking of the Latin American institutional framework

as exogenous and ﬁxed by rigid cultural bedrock. For example, it offers

an alternative, less culturally deterministic way of understanding the colo-

nial legacy in Latin America, including some of the most prominent ideas

of social embeddedness, such as Claudio V

eliz’s proposition that Spanish

America has been constrained economically by a centralist bureaucratic tra-

dition, or Morse’s notion that it mysteriously remain shackled to preindus-

trial Hapsburg institutions.

But although offering an alternative explana-

tion of long-run institutional rigidities, the self-reinforcement model does

not capture the complexity of institutional evolution in Latin America.

For example, it does not account well for the diversity of possible political

outcomes (multiple equilibria) or the inherent political instability that we

observe in the national histories. Though it gives a convincing explanation

of persistent wealth disparities, its predictions of political outcomes are too

deterministic. It oversimpliﬁes the political bargain among the elite, over-

states the elite’s political cohesiveness, and overlooks factors that caused

bitter political divisions and revolutionary political changes. Though long-

run institutional changes have characteristically followed slow, incremen-

tal paths, they have also, at critical moments, taken surprising turns with

unexpected consequences. This point stands out in the problems of national

reconstruction at independence.

INDEPENDENCE

Independence was a critical moment that opened up the future path of insti-

tutional change to many possibilities. When Latin American revolutionaries

Claudio V

eliz, The Centralist Tradition in Latin America (Princeton, NJ, 1980); Robert M. Morse,

“Heritage.”

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

174 Alan Dye

seized power, there was a real opportunity for national uniﬁcation and gov-

ernments based on liberal institutions, to which some leaders aspired; but

it did not materialize. As John Coatsworth writes, the opportunity for

modernization was “squandered.”

While trying to build the fundamental

institutions of government, liberals and conservatives collided into civil war

and led their countries into persistent political instability. The consequence

was a devastating economic collapse, which showed no recovery before the

reemergence of liberal reforms after mid-century.

The opportunity that independence offered for embarking upon a dif-

ferent growth trajectory reﬂects a larger principle – crisis often provides

a window of opportunity for altering the path of institutional change.

Yet discontinuous or revolutionary change does entail risk. The path of

institutional change in Latin America at independence could have taken

many directions. To understand the process of institutional change, one

need ask: What directions, other than the observed one, were reasonably

plausible? How feasible was a smooth transition to liberal government,

and what shapes could it have taken? The elite leadership reﬂected both

competing political objectives and fundamental uncertainty about the best

strategies for achieving them. One strategy called for cooperation, as in the

self-reinforcement model of political collusion. Another called for the dis-

mantling and replacement of the costly mercantilist institutions of the old

regime. Those hurt by imperial policies collided with those who feared the

social consequences of a radical deviation from known institutions. Polit-

ical demands for a liberal–conservative compromise did not materialize

because grounds for it were not found or were considered too costly. Politi-

cal bargaining to supply formal institutional changes was obstructed by the

break with former political institutions. Independence, which transferred

the seat of political authority to the colonies, coincided with the collapse

of the state, in effect leaving Latin American elites without functioning

institutions for making collective decisions. The ﬁrst task of recovery was

to reestablish formal political institutions to give orderly governance to

collective decisions–anecessary condition for an orderly political answer

to demands for economic reforms. But the implementation and enforce-

ment of reforms of economic policy depended on institutions higher up in

the legal and political hierarchy. Functioning constitutional, statutory, and

John H. Coatsworth, “Notes on the Comparative Economic History of Latin America and the United

States,” in Walter L. Bernecker and Hans Werner Tobler, eds., Development and Underdevelopment

in America: Contrasts of Economic Growth in North and Latin America in Historical Perspective (New

York, 1993).

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

The Institutional Framework 175

procedural laws, and the organizational capacity to enforce them, were a

necessary underpinning to effective policy reforms.

We are only beginning to understand the effects of postrevolutionary

political disorder on long-run economic growth. Most conventional inter-

pretations acknowledge that it introduced delays, but otherwise ignore it.

Latin America’s chronic economic disappointments are explained, instead,

by late entry into the world system or by the colonial legacy. But one cannot

properly assess alleged causes of underperformance without clearly spec-

ifying feasible counterfactual alternatives that might have given superior

performance, and it is improper to gauge actual outcomes against unreal-

istic alternatives. To comprehend the economic consequences of the insti-

tutional choices that led to political breakdown after independence, we

need to identify, describe, and examine how an alternative superior set of

institutions might have been implemented. The next sections look into

this question, beginning with the institutional foundations that created the

platform for supplying other institutional changes.

CONSTITUTIONAL FOUNDATIONS

Why did elites have so much trouble coming to an agreement over fun-

damental political institutions? Disagreement began with the question of

how to frame the new constitutions. Jeremy Adelman, in Republic of Cap-

ital, gives an illuminating account of how this controversy played out in

the birth of the Argentine nation.

Yet, generally, scholarly treatment of

the historical signiﬁcance of constitutions shows an inconsistency. Many

dismiss poorly enforced constitutions as meaningless and do not factor in

the consequences of constitutional ineffectiveness on the implementation

and enforcement of economic policy. Studies that focus on policy-level

institutional changes often inconsistently assume an underlying stable and

functional constitutional order.

Despite the inconsistencies, the conceptual link between political insti-

tutions and economic performance is straightforward. The appropriability

of private returns to investment was the main channel through which con-

stitutions and other political institutions affected private economic activity.

Individual decisions about how to hold and use assets were sensitive to the

Jeremy Adelman, Republic of Capital. Buenos Aires and the Legal Transformation of the Atlantic World

(Stanford, CA, 1999).

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

176 Alan Dye

existence of stable, predictable rules governing property rights, enforce-

ment of contracts, and other regulations. One can distinguish three differ-

ent ways that property-rights infringements discouraged investment. First,

weak governments sometimes failed to defend property rights from intru-

sion or seizures by other private individuals. To compensate, investors spent

larger sums on private security. Second, property holders had to confront

the “dilemma of the sovereign.” Governments sometimes conﬁscated pri-

vate wealth or selectively enforced the property rights of political supporters

or allies but not those of opponents. Third, investors had to consider the

political risks of both current and future governments. Even if current prop-

erty rights were enforced, investors had to factor in the political risks of

future governments. Without effective long-term constraints, such as con-

stitutional limits, a successor government could adversely change the rules

and alter the returns to an investment. Constitutions were the most funda-

mental layer of institutions provided to enforce property rights and private

contracts because they determined the rules for shaping the institutions of

enforcement in the hierarchy below them. However, if constitutional rules

were too variable or unpredictable, those uncertainties were also transmitted

to the lower-order institutions.

How stable have constitutional rules been in Latin American history?

Table 5.1, which displays the constitutions written between 1810 and 2000,

gives a striking image of the high frequency with which constitutional rules

have been altered and reﬂects a record of uncertainty about constitutional

rules and their expected longevity. Before reading too much into the table,

caution is warranted that the political charters called “constitutions” were

heterogeneous in character and intent. How or whether they were ratiﬁed

varied, and many were contested as invalid by political opponents. Others

were honored more in breach than in fact. But even ignoring the question of

enforcement, historians trying to count them must make judgments about

questionable cases; consequently, lists have varied depending on how these

judgments were made. Table 5.1 follows a predetermined set of guidelines.

It lists the dates of promulgation of newly framed constitutions, provisional

or incomplete charters that aimed to deﬁne bases of fundamental law or

governmental structure, and restorations of previous constitutions. It does

not include formal suspensions of constitutions. Though some dictatorial

regimes formally suspended the incumbent constitution, counting formal

suspensions would not capture most constitutional breaches. It was more

common for a dictatorial regime to ratify or decree its own constitution,

engage in selective violation of an existing constitution, or make liberal use

of constitutional emergency powers, known as “states of siege.” The list in

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

The Institutional Framework 177

Table 5.1 also understates the extent to which constitutional rules changed

by ignoring constitutional amendments and reforms, which, though rele-

vant, would be more difﬁcult to enumerate.

Granted these qualiﬁcations, two observations stand out. First, even

ignoring formal and de facto suspensions, the average duration of constitu-

tional documents was about ﬁfteen years. Omitting the 1810–29 period of

revolution and postrevolutionary transition to eliminate the bias of postrev-

olutionary provisional constitutions, the average duration was still only

twenty years. Yet, there was considerable variation in the degree of instabil-

ity. Between 1830 and 2000, the country with the longest average duration

was Chile, at seventy-four years, and the country with the shortest average

duration was the Dominican Republic, at six years.

the commitment problem

Liberal constitutional framers warned that their nations were at a crossroads

and that history offered two possible outcomes. They could follow the path

of the United States into a peaceful, prosperous transition to independent

rule, or they could follow the road of France, of failed constitutionalism and

political disorder. Peaceful and orderly transition depended on the coop-

eration of rival political factions. They, like scholars after them, adopted

a contractarian interpretation – the constitution was a multilateral pact or

contract, and the framers’ central problem was to negotiate a set of rules

that political contenders could universally accept. This was the function of

the constitutional congresses.

Recent analyses question the completeness of the contractarian inter-

pretation, noting that third-party enforcement could not ensure that the

new constitutions would be honored. Constitutional stability depended on

whether the ruling party or faction could credibly commit to honor the

constitution. This required that the ruling faction and its rivals each had to

perceive compliance as in their self-interest. But why would any rival strong-

man have surrendered his private militia and submitted to constitutional

limits if he perceived them as against his interests? Why would a faction that

ruled by force submit to a constitutional requirement that it step down to

allow an elected rival assume power? Rivals seeking to control the govern-

ment had incentives to comply with a constitution only if they expected to

beneﬁt in the long run from the orderly political process it would create. In

turn, those expected long-run beneﬁts depended on whether they believed

future ruling factions would also willingly accept the same limits. Few

countries had histories that offered any expectation that the constitution

would not be violated, overturned, or opportunistically altered within a few

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

Table 5.1. Constitutions of Latin America

Number of new constitutions

Years of most

All 1810–1829 1830–1916 1917–2000 durable const. Dates of promulgation of constitutions

Cuba 83532

1897, 1898, 1902, 1934, 1940, 1952, 1959, 1976

Brazil 81 1 6 67 1824, 1891,

1934, 1937, 1946, 1967, 1969, 1988

Uruguay 60 1 5 88 1830, 1918,

1934, 1942, 1952, 1967

Chile 96 1 2 92

1812, 1814, 1818, 1822, 1823, 1828, 1833, 1925, 1980

Argentina 96 1 2 96

1815, 1816, 1817, 1819, 1825, 1826, 1853, 1949, 1955

∗

Mexico 93 5 1 83 1812, 1814, 1824, 1835, 1836, 1843, 1847, 1857, 1917

Paraguay 61 23 70 1813, 1844, 1870, 1940, 1967, 1992

Colombia 14 4 9 1 105

1811, 1819

†

, 1821

†

, 1828, 1830, 1831, 1832, 1843, 1853, 1858, 1861, 1863,

1886, 1991

Costa Rica 12 2 8 2 61

1821, 1824

†

, 1841, 1844, 1847, 1848, 1859, 1869, 1871, 1882, 1943,

1949

Bolivia 15 1 9 6 60

1826, 1831, 1834, 1839, 1843, 1851, 1861, 1868, 1871, 1878, 1938,

1945, 1947, 1967, 1995

El Salvador 15 1 8 6 53 1824

†

, 1841, 1864, 1871, 1872, 1880, 1883, 1885, 1886, 1939, 1945

∗

1950, 1962,

1982, 1983

Peru 17 3 10 4 46

1823, 1826, 1828, 1834, 1837, 1839, 1855, 1856, 1860, 1867, 1868

1879, 1881

, 1920, 1933, 1979, 1993

Guatemala 91 35 42 1824

†

, 1839, 1851, 1879, 1921, 1945, 1956, 1965, 1985

178

Cambridge Histories Online © Cambridge University Press, 2008

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

Venezuela

26 3 12 11

38 1811,

1819

†

1821

†

, 1830,

1857, 1858,

1864,

1874, 1881,

1891, 1893,

1901, 1904,

1909, 1914, 1922,

1925, 1928,

1929, 1931, 1936,

1946, 1947,

1953, 1961, 1999

Nicaragua

12 1 6 5

1824

∗

, 1838,

1854, 1858,

1893,

1905, 1911,

1939, 1948,

1950, 1974,

1987

Dominican

Republic

17 9

28 1844,

1854, 1858,

1866, 1868, 1872,

1874, 1875,

1877, 1878,

1879, 1880,

1881, 1887, 1896,

1907, 1908,

1924, 1927, 1934,

1942, 1947,

1961, 1963,

1966, 1994

Ecuador

21 2 12 7

1812, 1821

†

, 1830, 1835,

1843, 1845,

1851, 1852, 1861,

1869, 1878,

1884,

1897, 1906,

1929,

1945, 1946,

1967, 1978,

1979, 1998

Honduras

17 1 10 6

21 1824

†

1839,

1848, 1865,

1873, 1874, 1880,

1894, 1904,

1906,

1908

∗

1921, 1924,

1936,

1957, 1965,

1982

Average

13.32

.06

.64

.757

Std Deviation

6.21

.84

.72

.726

Notes: Scholarly lists of Latin American constitutions, though similar

, generally do not agree in all details. The table lists promulgation dates of

constitutions, including federations and restorations of former constitutions.

Amendments to and reforms of an existing constitution are not

included. Dictatorial pronouncements of new constitutional documents

are included, but acts of suspension without replacement of the existing

constitution are excluded.

Legend: Shaded areas indicate constitutions that lasted no more than

10 years. Bold fonts indicate constitutions that lasted

50 years or more.

Italics indicate constitutions designated as provisional.

∗

indicates a restoration of a previous constitution.

†

indicates a constitution of a subsequently unsuccessful confederation.

Sources: See bibliographical essay.

179

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

180 Alan Dye

years of its ratiﬁcation. The credibility of commitment thus broke down

with distrust or doubts about its survivability.

The chronic difﬁculties of establishing effective constitutional con-

straints in Latin America thus reﬂect the general problem of creating a

credible commitment to limited government. Independence was a criti-

cal time because the collective decisions made then determined whether

subsequent governments encountered the commitment problem from a

prior condition of political instability or of constitutional order. The latter

could be maintained by the cooperation of rival political factions, but the

former would have to alter the historically determined distrust in the self-

enforcement of formal political rules. In a recent study, Douglass North,

William Summerhill, and Barry Weingast argue that the problem of estab-

lishing constitutional limits in Latin America was rooted in the institutional

discontinuities the newly created nations faced trying to reestablish formal

political institutions. Comparison with the former British North American

colonies brings the point into greater relief. Unlike the latter, where the

former colonies drafted their ﬁrst constitutions out of experience with rep-

resentative colonial assemblies, Latin America had more limited experience

with representative government (especially the Spanish colonies), and it was

at too narrow a scope to serve as an effective precedent for smooth tran-

sition to representative national governments. Furthermore, whereas the

states that emerged from British colonial America reduced the negotiation

costs of framing commitments to the federal constitution by deferring most

constitutional powers to the states, in Latin America, no comparable for-

mula was available to reduce the costs of negotiating a national constitution

that was acceptable to both liberal and conservative factions.

The ethnic and socioeconomic heterogeneity of Latin American popula-

tions seems to have been a critical factor. North, Summerhill, and Weingast

argue that the credible commitments to the state and federal constitutions

of the United States were underpinned by an ideology of constitutional

inviolability – a shared popular belief that a serious constitutional trans-

gression must be and would be met with popular resistance, if necessary.

If these were the underpinnings of limited government in the early United

States, then relative wealth disparities may be key to explain the difference

between the United States and Latin American experiences. Greater wealth

Douglass C. North, William Summerhill, and Barry R. Weingast, “Order, Disorder and Economic

Change: Latin American versus North America,” in Bruce Bueno de Mesquita and Hilton L. Root,

eds., Governing for Prosperity (New Haven, CT, 2000), 59–84.

P1: JzG

0521812909c05 CB929-Bulmer 052181290 9 October 3, 2005 8:53

The Institutional Framework 181

disparities in Latin America created sharp class distinctions, which caused

ideologies about constitutional inviolability to diverge. Uprisings against

perceived transgressions of government occurred, but elite and popular,

and liberal and conservative, views differed over what constituted a serious

transgression or which transgressions merited resistance. Elites perceived a

trade-off between two political risks – threats to property by government

predation or by social disturbance and popular invasions. A popular upris-

ing could result in a demand for redistribution rather than a defense of the

constitutional protections of property or political rights.

Historians have shown that leading spokespersons in the Latin American

constitutional assemblies were aware of these trade-offs but differed about

which threat weighed heaviest. Many conservatives and liberals feared pop-

ular uprisings more than tyranny and called for order and stability more

than political freedoms.

In the ﬁrst series of constitutional assemblies, dis-

trust of popular representation prevailed over cries for political openness

and resulted in suffrage rights that severely restricted popular political par-

ticipation. Although the evidence collected to date is sketchy, it shows that

around the middle of the nineteenth century, suffrage restrictions limited

voting rights to about 2 percent or less of the male population, compared

with about 24 percent in the United States (or 28%ofwhite males). By

1880, qualiﬁed voters in Latin American countries represented no more

than 4 to 6 percent of the male population, compared with 37 percent

(or 43%ofwhite males) in the United States.

Even so, the low rates of

suffrage actually overstate the effective political openness because electoral

fraud was widespread.

Early constitutional formulas showed concerns for civil protections

against tyranny but, over time, those concerns were given less weight and

constitutions were revised to strengthen presidents and reduce constitu-

tional checks to their power. An important device, included in almost all

constitutions, was the state of siege, which extended to presidents’ extraor-

dinary powers to suspend constitutional rights in the event of a social distur-

bance or similar threat. These powers contributed to a lack of moderation

in political competition. States of siege were used liberally to suppress polit-

ical opposition, rig electoral outcomes, and justify opportunistic violations

See especially Jeremy Adelman, Republic of Capital ; Charles A. Hale, Mexican Liberalism in the Age

of Mora, 1821–1853 (New Haven, CT, 1968); and Tortcuato S. Di Tella, National Popular Politics in

EarlyIndependent Mexico, 1820–1847 (Albuquerque, NM, 1996).

Stanley L. Engerman and Kenneth L. Sokoloff, “The Evolution of Suffrage Institutions in the New

World” (National Bureau of Economic Research Working Paper 8512,October 2001).