WPS3859



                EVALUATING RECIPES FOR DEVELOPMENT SUCCESS*




                                               Avinash Dixit

                                           Princeton University




World Bank Policy Research Working Paper 3859, March 2006


The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the
exchange of ideas about development issues. An objective of the series is to get the findings out quickly,
even if the presentations are less than fully polished. The papers carry the names of the authors and should
be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely
those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors,
or the countries they represent. Policy Research Working Papers are available online at
http://econ.worldbank.org.




*This is a slightly revised version of a paper presented at the World Bank DEC Lectures on April 21, 2005.
A more thorough revision is planned; therefore feedback is still welcome. For many valuable comments
and suggestions already received, I thank Philip Keefer (my discussant at the conference), Daron
Acemoglu, Karla Hoff, Boris Pleskovic, and Dani Rodrik. But I must emphasize that they are not
responsible for my interpretations and choices; in fact they will disagree with much if not all of my
interpretation of the literature.


                                                       1

The main purpose of most scholarly research, both theoretical and empirical, is to

improve our understanding of the phenomena and processes being studied. The work may

lead to useful prescriptions for policy, but that is usually the end-point of a long and

winding path. Often, however, especially in matters of economic development and

growth, the problems under investigation are large and urgent, so practitioners want

immediate answers. Academic researchers are also tempted to offer suggestions for

policy, however tentative or incomplete.


        Work on institutions and growth is a case in point. When I present my own

theoretical models to audiences, I am almost invariably asked to interpret the results for

their policy implications. The listeners usually find my responses meager and

unsatisfactory, and I am compelled to agree. Today I want to take this opportunity to

apply the same, admittedly very stringent, test of policy relevance to all academic

research in this area. I will interpret all findings of this body of work as if they were

prescriptions for policy or institutional reform, and question the validity and practicality

of these interpretations.


        I am deliberately going to be provocative and critical, but will try to be

evenhandedly so. I hope that my remarks will give everyone some incentive to think

further and harder. I also hope to help scholarly researchers better see their own work in

the context of the bigger picture, and help practitioners better appreciate the difficulties of

drawing implications for action from an ongoing process of academic exchange.


        My primary focus is on institutions of economic governance ­ protection of

property rights, enforcement of voluntary contracts, and provision of public infrastructure

and services that support private economic activity. This relates on one hand to the

broader question of the emergence and reform of deeper institutions such as democracy,

in so far as that influences the shape and function of the institutions of economic

governance and in turn affects economic performance. It also relates on the other hand to

some specific organizations that implement the rules of the game put in place by the

institutional setting, and even to some specific policies in so far as they illustrate the



                                                 2

issues involved in thinking about institutional change. There is a spectrum from deep

institutions to specific policies; my discussion is located somewhere in the middle of this

spectrum, with focus blurring toward either end. One more restriction, dictated by the

limitations of my own expertise, is that I focus on microeconomic issues and institutions,

not those of macroeconomic policy.


        Even within these confines, the literature is huge, and I cannot hope to include

even a substantial fraction of it in the space and time available. However, those left out

should be relieved, not upset: the policy recipes that emerge from this literature are

almost invariably unsatisfactory, so being omitted from the list and the implied criticism,

however friendly and constructive, may perhaps be regarded as a good thing.


        Critical assessments of the contribution of academic research to development

policy have a long tradition. Most recently, Easterly (2001, chapters 2-7) has produced

such a comprehensive, compelling, and entertaining critique of theories of economic

growth. He concludes that the prescriptions of these theories ­ policies to increase saving,

investment, foreign aid, and education, to modernize technology, and to reduce

population growth and debt ­ mostly failed because other policies of LDC governments

often maintained or created perverse incentives. He concludes that the path to success

must begin by setting up "quality institutions" (p. 252), but says little about how this can

be done. I will try to extend the critique to the research on institutions, by subjecting this

literature to a prospective test of usefulness; a retrospective test will have to wait for the

outcomes of any attempts to use it.


        Here are some examples of broad categories of ways in which the accumulated

research on the role of institutions in development stops short of giving any useful or

reliable policy prescriptions. (1) Some of the work finds that development success is

contingent upon some historical or geographic preconditions that most countries do not

meet. (2) On many crucial issues, from deep institutional divides of democracy versus

authoritarianism, to general policy approaches such as comprehensive change versus

gradualism and sequencing of reforms, all the way to very specific policies toward



                                               3

population, inflation, regulation, and so on, the literature includes support for

diametrically opposed prescriptions, each of which sounds quite plausible when

presented on its own. Thus, unlike the theories of economic growth that serially offered

definite answers (which did not work in practice), theories of institutions simultaneously

offer different and conflicting answers. (3) The recipes for success are often motivated

by ideological fashions or a herding instinct that follows the latest perceived success. (4)

The recipes that seem most promising are very often at a level of generality that stops far

short of practicability. Examples are advice to build institutions that complement existing

ones, or to build participatory democracy; this leaves out all details of how to identify the

complementarities, or how to convert an authoritarian regime into a participatory

democracy.


        My conclusion from this overview of the literature is that it has not elucidated the

mysteries of successful development, and we cannot draw many reliable and practicable

policy conclusions from it. But I do not want to be entirely negative. In the concluding

section I develop a promising framework or methodology of research that combines

general conceptual and empirical findings from academic research and the experience of

practitioners, to help us narrow down or identify the causes of failures in the cases of

individual countries. This can then serve as the first step toward remedying the failures or

removing the obstacles.

1. Infeasibility ­ The "Irish" Recipes for Success


        Numerous researchers have found significant correlations across countries

between various geological, ecological, geographic, and historical data and economic

performance. Taken literally, these findings constitute a message of pessimistic

determinism: if your country lacks the right prior or starting conditions, its economic

future is bleak. Of course many of the correlations are not to be taken literally, but I will

start by listing them all on an equal footing of literal interpretation, and then develop

further arguments and different interpretations. To make them amusing and memorable, I

call these findings "Irish recipes for success", after the story of the Irishman who was

asked for directions and replied "If I wanted to go there, I wouldn't start here."



                                              4

        Many geological and geographic characteristics have the greatest appearance of

determinism. The extremely long run processes described by Diamond (1997) are of

course the most extreme examples of such determinants of economic outcomes: an east-

west rather than a north-south alignment, the kinds of animals and plants that were native

to the region, and so on. We are also told that it helps to be far from the equator, in the

temperate rather than the tropical belt (Gallup, Sachs, and Mellinger 1999, Hall and Jones

1999), unless your country is a small Pacific island, in which case distance from the

equator is one of the factors conducive to a total collapse of your ecology, economy,

civilization, and population (Diamond, 2004). Being far from the core areas of the world

economy is a major piece of bad luck, and being landlocked makes it worse (Gallup,

Sachs, and Mellinger 1999). You might have thought that your country was fortunate to

be endowed with abundant natural resources, but the facts are just the opposite. Resource

abundance worsens your growth prospects; Sachs and Warner (2001) summarize and

extend these findings; see also Engerman and Sokoloff (1997). If your land is suitable for

plantation agriculture, that is also bad for your economic performance (Engerman and

Sokoloff (1997).


        On the other side of the balance, if your region was sparsely populated and poor at

the time of the spread of European colonialism in the 1500s, your country is more likely

to be economically successful now (Acemoglu, Johnson, and Robinson 2002, Engerman

and Sokoloff 2002). It was a special help to be colonized by the British, so a larger

fraction of your population became fluent in English (Hall and Jones, 1999), and inherit

British rather than French or other legal and administrative institutions, giving better

protection to the rights of small shareholders and therefore allowing broader and more

efficient capital markets (La Porta et al., 1998, 1999). If the local diseases were more

likely to take heavy toll of the early colonizers, your economic performance today is less

likely to be good (Acemoglu, Johnson, and Robinson 2001). And numerous economic

and social scientists from Weber onward have appealed to higher authority, finding roots

of success in Protestantism, Confucianism, and so on.




                                               5

        Interpreted literally as recipes or policy recommendations, these require a less-

developed country to use plate tectonics to move itself to a more favorable location, or to

turn the clock back and invite British colonizers, of course cleaning up the local disease

environment and getting rid of mineral resources beforehand. As a practical matter, these

findings are merely telling countries to accept their fate. Perhaps at best they could be

reinterpreted as telling the fortunate countries to share their good luck by giving massive

amounts of aid to the unfortunate.


        Of course this is not the interpretation the researchers intend; they intend many of

their history and even geography variables to have only indirect effects on economic

outcomes via some other proximate determinant of success, or to be mere econometric

instruments used for identifying the direction of causation. This is emphasized by the

authors of these articles, and by others such as Rodrik (2004 a), who also makes an

important cautionary remark: "Finding an appropriate econometric instrument is not the

same as finding an adequate explanation." The supposed proximate determinants of

success are categorized as institutions (Acemoglu, Johnson, and Robinson 2001, 2002,

Engerman and Sokoloff 1997, 2002), or social infrastructure (Hall and Jones 1999). Here

are some examples of the suggested channels.


        If a colony had rich mineral resources or climatic and soil conditions conducive to

plantation agriculture, and a large or dense population that could be press-ganged into

work in such activities, then the European colonizers established institutions such as

slavery and inequality that facilitated extractive and plantation activities. Institutional

change is a very slow process; old institutions persist or shape today's institutions. But

they are less suitable for modern economic conditions, where broader participation in the

decision-making process, education, access to economic opportunities, and protection of

everyone's property rights are needed for success. Conversely, areas that were sparsely

populated and poor in the 1500s were likely to lack such get-rich-quick opportunities.

There the European colonizers set up institutions conducive to longer-term economic

success. (See Engerman and Sokoloff 1997, 2002; Acemoglu, Johnson, and Robinson

2002. See also an excellent account of this research in Hoff 2003.)



                                               6

        If European colonizers were subject to local diseases and suffered high mortality

in a colony, they were likely to establish extractive economies and institutions; if they

had good prospects of long lives as producers, they were likely to settle, and establish

economic activities and supporting institutions conducive to long-run economic success.

(See Acemoglu, Johnson, and Robinson 2001; see also the critique by Albouy 2004, and

reply by Acemoglu, Johnson, and Robinson 2005.)


        More broadly, Hall and Jones (1999) regard "various correlates of the extent of

Western European influence," primarily distance from the equator and the prevalance of

English, French, German, Portuguese, and Spanish as first languages in the countries, as

instruments for "social infrastructure," measured by an index of lack of corruption

(government antidiversion policies) and openness to international trade.


        The need for instruments arises because of potential reverse causation: good

institutions are likely to be conducive to good economic performance, but citizens of

economically successful countries may also demand and implement good instiutions.

However, Kaufmann and Kraay (2002) find a weak but negative reverse causation,

suggesting the absence of a virtuous circle between better governance and better

economic outcomes. Keefer (2004 b) also argues for a weak or even negative reverse

causation, combining conceptual arguments and empirical evidence. But even negative

reverse causation can create an econometric problem requiring instrumental variables.


        Formally, let yi denote the economic performance of country i, let xi be a vector of

measures of its institutions, and zi a vector of other relevant variables. The hypothesis

being tested is


(1)     yi = a1 + a2 xi + a3 zi + ei


There is reverse causation from the yi to the xi, so one invokes instrumental variables that

are assumed to affect the xi, but have no direct effect on the yi. Thus



                                               7

(2)     xi = b1 + b2 yi + b3 zi + b4 gi + ui


One then estimates (1) by two-stage least squares, estimating the xi from the first-stage

equation


(3)     xi = c1 + c2 zi + c3 gi + vi


and using the estimated values in the regression for (1). This requires that the gi do not

appear directly in (1), and are uncorrelated with the error term ei in that equation. One

constructs various stories for why that should be so. For example, Hall and Jones (1999)

say the following about their "European influence" instruments: "we must ask whether

European influence was somehow more intensively targeted toward regions of the world

that are more likely to have high output per worker today. In fact, this does not seem to

be the case. On the one hand, Europeans did seek to conquer and exploit areas of the

world that were rich in natural resources such as gold and silver or that could provide

valuable trade in commodities such as sugar and molasses. There is no tendency for today

for these areas to have high output per worker. On the other hand, European influence

was much stronger in areas of the world that were sparsely settled at the beginning of the

sixteenth century ... Presumably these regions were sparsely settled at that time because

the land was not especially productive given the technologies of the fifteenth century. For

these reasons, it seems reasonable to assume that our measures of the Western European

influence are uncorrelated with [the error term]." This discussion also relates to the

instruments and their discussion in the other work of this kind cited above.


        Many of these stories do sound plausible. But how does one formally test them?

In other words, how does one test the difference between the system (1)-(2) and an

equation like


(4)     yi = d1 + d2 zi + d3 gi + wi




                                              8

that allows economic success to depend directly on the supposed instruments and not on

the supposed proximate institutional determinants at all?


        If there is just one instrument, for example colonial mortality in the primary

specification of Acemoglu, Johnson, and Robinson (2001), then the system (1)-(2) is just-

identified, and econometrically indistinguishable from (3)-(4), unless one is willing to

make other assumptions about error variances or measurement errors; this is the

procedure adopted by Hall and Jones (1999), Acemoglu, Johnson, and Robinson (2001),

and Kaufman and Kraay (2002). This involves some other untested assumption. If there

is more than one instrument (gi is a vector), then one can perform overidentification tests,

but these are weak and too likely to accept the null hypothesis that the gi are excluded

from (1). In Acemoglu, Johnson, and Robinson (2002), the geographic reversal of fortune

instrument is tested in such a way, assuming that colonial mortality is a genuine valid

instrument; but that is not problem-free as was just argued.


        My view is that the idea that geographic and historical variables are merely

instruments for institutional determinants of economic success is supported more by the

intuitive appeal of the stories told than by the statistical significance of the tests

performed. The value of rhetoric should not be ignored, but I wish the econometric

evidence were stronger and more compelling. This is an area where more and sharper

research is badly needed.


        But let us proceed on the basis that we are convinced by these stories, and accept

that institutions are important. What policy implications follow? The literature reviewed

above gives us no guidance on this question; whether geography or history have a direct

effect or an effect via institutions, a recommendation to change one's geography or

history is useless. One must forget about history and geography and try to implement the

relevant institutions directly. For advice on that, we must turn to some other research.


2. Contradictions ­ "Doctor Dolittle" Recipe Pairs




                                               9

        There is no shortage of academic research that identifies institutions and policies

to promote economic growth and development, and proffers advice on how to acquire

good institutions. Indeed, "Institutions for High-Quality Growth: What They Are and

How to Acquire Them" is the title of Rodrik's (2000) paper. The trouble is that for every

paper that endorses one kind of institution or policy, one finds another that makes

precisely the opposite claim. Each is written by a prominent economist, and contains

impressive arguments and evidence to support the recommendation being made.


        I will now examine several dimensions of the contradictory conclusions of the

literature on institutions and policies. They range from the largest choices of the deepest

institutions that govern the society and the polity, to some minutiae of specific policies.

Once again I have chosen a whimsical and memorable label for this literature. To remind

us of the two-headed animal "pushme-pullyou" in the book and musical, I call these the

"Doctor Dolittle" pairs of recipes for development.


A. Democracy versus Authoritarianism


        At a broad and general level, cross-country regressions yield mixed results on the

question of whether democratic or authoritarian governments are better for growth. For

example, Barro (1999, p. 61) found a relatively poor fit and an inverse U-shaped

relationship. He suggested that "more democracy raises growth when political freedoms

are weak, but depresses growth when a moderate amount of freedom is already

established." Persson (2005), using cross-sectional as well as panel data, finds that the

crude distinction between democratic and non-democratic forms of government is not

enough; the precise form of democracy matters for policy design and economic

outcomes. "Reforms of authoritarian regimes into parliamentary, proportional, and

permanent democracies seem to foster the adoption of more growth-promoting structural

policies, whereas reforms into presidential, majoritarian, and temporary democracy do

not." However, Keefer (2004 a), after surveying a wide-ranging literature on electoral

rules and legislative organizations, concludes that they affect policies but are not a crucial

determinant of success: "electoral rules ... almost surely do not explain why some



                                              10

countries grow and others do not," and "the mere fact that developing countries are more

likely to have presidential forms of government is unlikely to be a key factor to explain

slow development."


          What are the reasons or mechanisms that can explain the claimed superiority of

democracy? Rodrik (2000) emphasizes the importance of local knowledge in the process

of successful institution-building, and argues that participatory democracy is meta-

institution that facilitates such use of local knowledge and thereby enables higher-quality

growth. Besley, Persson, and Strum (2005) using data from the United States argue that

political competition is a key to better economic policies and outcomes, and this also has

intuitive appeal. Besley and Burgess (2002) using panel data from India find that an

informed and active electorate leads to effective incentives for governments to respond to

economic problems, and that mass media play and important part in this process. In other

words, democracy succeeds by facilitating voice and participation. Bardhan (2005, ch. 1)

also stresses the importance of democratic participation in a generalized interpretation of

the "rule of law." Islam (2003) emphasizes the role of information and transparency in

improving the quality of governance institutions.


          Some who favor democracy over authoritarianism in a comparison of ongoing

systems may nevertheless be concerned about costs of disruption in a transition to

democracy. For them, Rodrik and Wacziarg (2004) bring good news: using annual panel

data for over 150 countries (with separate subsamples analyzed for some questions), they

find that transitions from authoritarian to democratic regimes lead to striking and

statistically significant improvements in GDP growth per capita; for example, the effect

is as large as 2.8% per year for sub-Saharan Africa. Growth also becomes less volatile
                                                               1


after a transition to democracy. Declines in growth precede, not follow, such transitions.


1Actually, Rodrik and Wacziarg find that either kind of transition is followed by better economic growth,
but the effect is smaller for transitions from democracy to authoritarian rule than for transitions in the other
direction. This suggests an alternative hypothesis that can explain the same facts. Suppose that purely
exogenous random shocks are the main determinants of growth, but the public gives the rulers credit for
good results and blame for bad results. Rulers are overthrown when the results are particularly bad. Then
the average improvement in growth following a regime change is just a reversion to the mean, and the
effect is stronger for transitions to democracy because authoritarian rulers can be overthrown only when the
economy performs drastically badly whereas democratic governments may fall for smaller shortcomings.


                                                     11

         These are just a few examples from a vast literature, and they add up to a message

that is pleasing and even uplifting to many modern academics and policy practitioners

alike ­ democracy is good not only for its moral and human appeal, but also for its

economic performance. This would be a "warm glow recipe" or a "happy ending."


         But not so fast. There is an equally impressive emerging literature that makes a

serious case for authoritarian governments and institutions when it comes to starting

growth and development. Glaeser, La Porta, Lopez-de-Silanes and Shleifer (2004) argue

that less-developed countries that achieve economic success do so by pursuing good

policies, often under dictatorships, and only then do they democratize. While these

conclusions are controversial, these authors' criticisms of the measures of institutions

used in the research that argues for the primacy of institutions in general, and of

democracy in particular, are telling. Giavazzi and Tabellini (2004) find a positive

feedback between economic and political reform, but they also find that the sequence of

reforms matters, and countries that implement economic liberalization first and then

democratize do much better in most dimensions than those that follow the opposite route.


         This debate looks likely to go on for a long time, with contending arguments

about basic concepts and theories as well as empirical methodologies, data, and

anecdotes. Here I would like to say something brief about two questions.


         First a conceptual matter: what feature or features of policies are important for

good economic outcomes, regardless of what kind of government makes those policies?

Here much of the literature does find one point of agreement: credibility of commitments

is vital. Acemoglu (2003) argues that the fact that political contracts lack third-party

enforcement makes it harder to make credible commitments, and this explains the

absence of a Coase Theorem ensuring efficient outcomes in political bargaining. In his

comprehensive overview of different dimensions of governance and their effects on

economic development, Keefer (2004 a, b) also identifies the ability to make credible

commitments as crucial. Bardhan (2005, ch. 4) offers a good discussion of the concept of



                                               12

credibility, and of its lack as a barrier to reform. In their illuminating analytical narrative

of Argentine politics, Spiller and Tommasi (2003) that the institutions and organizations

of political bargaining in that country, and the volatility of its economic environment,

combined to inhibit the capacity of its policymakers to "undertake efficient intertemporal

political exchanges," and this led to "low-quality policies" on many issues such as

regulation, and control of inflation.


        How do democracy and authoritarianism compare in this respect? Keefer (2004 a)

stresses the role of checks and balances in achieving credible commitments, and these are

more likely to be present and effective in democracies. However, one might argue that

commitments coming from a dictator or authoritarian ruler should be more credible, if

anything, so long as he is stable in his rule. And such a ruler does have incentives to

pursue policies that enhance economic success. Even if his aims are predatory, if he is

sufficiently stable in his position, he will best increase his own take if the pie is larger;

this is Olson's (1993) stationary bandit. Whereas, if he fears being overthrown,

improving the people's economic well-being may be his best way of postponing that fate.


        China is usually cited as an example of an authoritarian regime where formal

property rights are not formally protected, but are de facto quite secure to the point that

the country can attract large investments from abroad. Different explanations of this can

be found in the literature. Qian (2003) and Rodrik (2004 b) find it in the Chinese

institutional innovation of township and village enterprises (TVEs). This "force[ed]

entrepreneurs into partnership with their most likely expropriators, the local state

authorities. ... Local governments were keen to ensure the prosperity of these enterprises

as their equity stake generated revenues directly for them." This puts the local authorities

in a role similar to Olson's stationary bandit. However, the strategy sounds very similar

to the concept of "insider privatization" pursued in Russia, and supported by Shleifer and

Treisman (2000, pp. 31-2), which did not work so well there. What was the difference?


        McMillan (2003, pp. 98-100) argues that in the phase of China's agricultural

reform, "productivity gains were achieved without formal legal recognition of farmers'



                                                13

ownership rights. ... Although the authorities are able to renege on contracts, they have

refrained from doing so with sufficient predictability that the farmers are motivated to be

productive." He finds the probable reason in the "specifics of time and place. ... The

communist government faced no challenge, [but] its legitimacy as the government, and

its ability to preempt any future political opposition, rested on its delivering economic

growth. High officials in Deng Xiaoping's government understood enough about

economics to recognize that growth requires markets and markets require assured

property rights. The Communist Party had retained its highly disciplined organization and

so was able to prevent self-seeking behavior by low-level officials." This is the idea of a

dictator operating under a constraint that captures his fear of losing power. Indeed, it

appears that China retains several aspects of democracy at lower and middle levels of

institutions and economic policymaking: there are some genuinely contested elections at

these levels, press criticism of officials at this level is tolerated and perhaps even

encouraged, and corruption is swiftly and severely punished when detected. Only at the

top level is the Communist Party's rule rigid and unchallengeable.                2




         Thus it may be possible to combine the two forms of government institutions for

optimal economic performance. However, there is always the problem of how to ensure

ex ante that your authoritarian ruler turns out to be a benevolent promoter of economic

success and not a predatory despot. And if your country has the misfortune of getting a

predatory despot, you face the bigger problem of how to get rid of him. Similarly, the

voluminous literature has almost unanimously made a convincing case that credible

protection of property rights is essential for investment and innovation and therefore an

important necessary condition for development success. But this literature gives us no

guidance on the operational question of what to do if one finds oneself in a policy role in




2One other feature of economic reforms in China is often highlighted and praised (for example Qian 2003,
Rodrik 2004 b): China preserved established entitlements in an inframarginal manner, while liberalizing
activity at the margin. For example, the state purchase system and price subsidies for agriculture were
retained, thereby protecting government revenue and urban living costs, but gave farming households the
freedom to sell the remaining output after meeting the state quotas at free market prices. However, this
policy decision is conceptually distinct from the issue in this section, namely democracy versus
authoritarianism.




                                                      14

a country where a despot cancels these rights at his whim, or a democratic government

does so at the behest of politically important special interests.


B. Formal versus Informal Governance Institutions


          Institutions to protect property rights and enforce contracts can be formal ­ the

state's apparatus of legislation, policing, and the judiciary ­ or informal ­ social

networks, communication channels, and norms, and for-profit information and

enforcement services including but are not limited to organized crime. Many variants and

mixtures of formal and informal systems exist in practice in most countries. We see

private adjudication that has the backup of formal enforcement; for example industry-

based arbitration panels can be recognized by courts, who then merely act to enforce the

judgments issues by arbitrators. Rubin (1994) recommended this approach for building

governance institutions in transition economies. We also see private enforcement of

formal laws, ranging from private initiation of litigation to private infliction of

punishments; Berglöf and Claessens (2004) and Hay and Shleifer (1998) favor various

forms of this approach.


          What are the relative merits or flaws of formal and informal institutions of

economic governance? The literature on this question also contains a bewildering and

often mutually contradictory spectrum of arguments and evidence.                     3




          A useful conceptual framework for studying the formal-informal distinction is the

distinction between relation-based and rule-based governance of contracts. In relation-

based governance, transactions occur between parties both of whom are members of the

same group or network, based on ties of language, ethnicity, locality, or the like.

Information circulates within the network, and helps each member both in the search for


3I restrict attention to institutions of property right protection and contract enforcement, but institutions and
agencies for regulation of product and labor markets, taxation, trade and so on show an equally bewildering
variety, and equally complex considerations as regards their relative merits. Islam (2004) considers the
institutions that affect international trade, and concludes that "increasing trade integration is consistent with
a wide array of institutional choices." However, it is less easy to specify which choice is best under what
circumstances.


                                                         15

a partner with whom to transact, and in obtaining information about the capability and

trustworthiness of a prospective partner. Contracts are personal and implicit agreements.

Any breaches of promises are communicated to the group, and other members of the

group can punish the miscreant by refusing to trade with him in the future, or by extra-

economic sanctions such as social ostracism or worse. Rule-based governance, by

contrast, relies on formal contracts and their enforcement by the courts, or the police if

necessary.


        Although this distinction was known and studied for many years in many

contexts, Li (2003) gave it sharper analytic content by clarifying the main difference in

the costs of establishing and operating the two systems. Relation-based governance has

few fixed costs. It does not require a framework of laws, regulation, or of course, of

courts; one just starts dealing with one's friends and neighbors. But its marginal costs are

significant and increasing as the scope of trade expands: one starts by dealing with close

neighbors whose trustworthiness is better-known, and who are more likely to behave well

in dealings with one because there are likely to be more frequent and larger future links

among close neighbors. As one's business expands, it becomes necessary to deal with

more distant strangers, who are less likely to be trustworthy. By contrast, rule-based

governance has high fixed costs of legislation, regulation (such as reporting and auditing

requirements that create information about traders), and enforcement. But once these are

in place, one can deal with strangers at low marginal cost. Because of these cost

differences, the relation-based system is better at small scales of transaction, and the rule-

based system at large scales. Dixit (2004, ch. 3) constructs a formal model to examine

and compare equilibria under the two systems; this elucidates the interaction of different

parameters and mechanisms that determine the limits of the relation-based system.


        This was precisely what Greif (1993, 1994, 1997) found when comparing two

groups of traders around the Mediterranean in the medieval period. The Maghribis relied

on relation-based governance within their tight group of Jewish traders; this worked well

up to a point but eventually could not cope with the expansion of trade to newer and

farther locations. The Genoese traders used a more formal system of registering contracts



                                              16

with the authorities in Genoa, and having any disputes investigated and adjudicated by

these authorities; this proved better as trade expanded. More recently, microcredit has

been successful in many small communities and groups, but expanding its scope to larger

financial markets for lending, borrowing, and investing has proved more problematic. A

possible approach is to link the small communities together at an upper tier using the kind

of system Bernstein (1992) describes for the diamond industry, or the "community

responsibility system" in medieval Europe studied by Greif (2004).


        Worsening quality of information and communication is not the only problem that

limits the expansion of informal institutions. Another issue is inherent in the concept of a

"community" that develops these informal institutions. This is a group of "insiders,"

which by its very nature requires a concomitant definition of "outsiders" who are to be

excluded. Therefore informal institutions usually erect entry barriers that inhibit fuller

exploitation of opportunities and development of capital markets. Fafchamps (2004, pp.

455-456) develops this idea and offers examples.


        When comparing the relative merits of formal and informal, or rule-based and

relation-based, institutions of governance, it may be important to distinguish between

enforcement of contracts and protection of property rights. Informal, relation-based

systems seem to be less successful in protecting property rights than in enforcing

contracts. Li (2003) confines most of his discussion to contracts, and in fact tacitly

assumes a background of formal institutions that include property right protection: he

says that relation-based governance "requires only minimal public order ­ that is, the

general absence of rampant robberies or confiscation" to get going. And Gambetta's

(1993) study of the Sicilian mafia indicates that their performance in enforcing contracts

could be socially beneficial, whereas in their role of protecting property, "protectors,

once enlisted, invariably overstay their welcome" (ibid, p. 198). Even with contracts,

information services about trustworthiness do not generate problems of violence, whereas

enforcement services to punish breaches do (ibid, p. 18); see also Dixit (2004, ch. 4).




                                              17

        De Soto (2000) emphasizes the importance of formal recognition of property

rights (pp. 5-6): "in Asia, Africa, the Middle East, and Latin America, ... most of the

poor already possess the assets they need to make a success of capitalism. ... But they

hold these resources in defective forms: houses built on land whose ownership rights are

not adequately recorded, unincorporated businesses with undefined liability, industries

located where financiers and investors cannot see them. Because the rights to these

possessions are not adequately documented, these assets cannot readily be turned into

capital, cannot be traded outside of narrow local circles where people know and trust

each other, cannot be used as collateral for a loan, and cannot be used as a share against

an investment." But he recognizes the importance of building upon local knowledge (De

Soto, 2004, p. 108): "As I strolled through rice fields [in Bali], I had no idea where the

property boundaries were. But the dogs knew. Every time I crossed from one farm to

another, a different dog barked. Those Indonesian dogs may have been ignorant of the

formal law, but they were positive about which assets their masters controlled." So

officials who wanted to set up a formal property system could "by traveling their city

streets and countryside and listening to the barking dogs, ... gradually work upward."


        Most theoretical and empirical studies find that purely informal institutions of

governance eventually act as a constraint on growth and expansion of trade. No country

switches to purely formal institutions; even in the most advanced countries, much

economic activity continues to be governed by relational and private ordering "under the

shadow of the law"; see Bernstein (1992, 2001) for some examples of such industry-

based adjudication in the U.S. But it remains important for less-developed countries and

former socialist countries to develop more formal institutions that can govern increasing

volumes of arm's length transactions. This transition brings its own difficulties, and

evidence as well as arguments on this issue yield mixed results. Some formal institutions

can meld well with the existing informal ones, for example Rubin (1994) and Dixit (2004

ch. 2) argue that arbitration does, although Widner (2000) has challenged the efficacy of

specialized courts. In other instances, the two institutions clash with each other and the

formal one may lose; Ensminger (1997) gives an example of land title reform in Kenya.

Traditional property rights to land in many parts of Africa are a complex system. Clan



                                             18

chiefs grant titles to individual families. Sales are subject to their approval and also that

of family heirs since all sons usually have expectations of equal division. Many family

stakeholders have usufruct rights. When the Kenyan government attempted to impose a

system of formal land titles, this ran into conflict with the traditional arrangements. The

expected capital market did not develop because lenders knew that foreclosure was

infeasible in the face of opposition from family and community, so the land could not be

used as collateral. Attempts to consolidate scattered holdings for scale economy reasons

did not work because there was a good economic reason (insurance) for the scattering.

Many formally registered titles are now being allowed to lapse and revert to older

arrangements, and the laws are being changed to resemble traditional forms more closely.

Finally, theoretical literature using a repeated game framework shows how a partial

improvement of an imperfect formal system, by providing a better outside alternative and

thereby making the consequences of breaking a relational contract less bad, can worsen

the outcomes from the existing informal system (Baker, Gibbons, and Murphy 1994,

Dixit 2004 ch. 2).


C. Comprehensive and Rapid, versus Sequential and Gradual, Reforms


        This is an old debate in the development field, and continues unabated.

Arguments for speed include the following: (1) Windows of opportunity for reform

present themselves rarely in political negotiation, so they should be seized when they do

arise. (2) Reforms cause some pain to some people or groups, and gradualism prolongs

this period of pain, thereby risking the whole reform. (3) Policy actions convey useful

signals about the government's intentions to financial markets and investors, and in the

standard Spencian manner, a reforming government must engage in excessive signaling

to separate itself from a less purposeful one. (4) The different dimensions of reform are

strategic complements, and results will be poor unless all dimensions are tackled jointly

(see Kramer 1993 a for the "O-ring" theory of this general class of phenomena). Others

favor a more cautious approach. Their arguments include the following: (1) Drastic

reforms, by inflicting severe pain, and likely to arouse more opposition and thereby

create a greater risk of political opposition leading to failure. (2) Institutional change is a



                                               19

slow process since it requires changes in long-held expectations. Prominent early

supporters of speed and comprehensiveness include slund (1995); those on the side of

caution include Murrell (1992) and Desai (1995). The World Bank (1996), in its World

Development Report, on balance emerged in favor of speed, but with many cautions and

caveats. Heybey and Murrell (1999) offer an empirical assessment based on more recent

data. They find that the initial conditions of a country are more important than policy

changes in determining its economic performance during the first few years of transition;

that is, speed is less important either way. They also find a negative feedback in reform:

higher levels of initial liberalization are likely to slow down subsequent reforms.


         A somewhat different recent argument in favor of one kind of gradualism, and a

new view of optimal sequencing, says that the most productive institutional changes in

the early phases of development are very small and easy to implement; the harder job is

to build on them and sustain growth beyond a medium level. Thus Pritchett (2003, p.

148) concludes from an analytical case study: "Under a regime that has reasonable

instiutional stability and is not completely dysfunctional, a rapidly increasing level of

GDP per capita is possible up to semi-industrialization. ... a rough and ready system for

enforcing contracts and providing for stability of investors' expectations" suffices to

"support quite rapid growth if beginning from a low base." Even some corruption can be

tolerated so long as it is "predictable," that is, investors are clear as to whom to bribe and

how much, and can rely on the official delivering on the quid pro quo. "However, what

trips countries up is the transition from one set of institutions to another," (emphasis in

the original) because this requires a change in expectations and creates uncertainty during

the transition. Rodrik (2003, p. 17), in his overview of Pritchett's paper and several other

others, similarly concludes that "the policies required to initiate a transition from a low-

income equilibrium to a state of rapid growth may be qualitatively different from those

required to reignite growth for a middle-income country." This fits with the ideas of Li

(2003) and Dixit (2004, ch. 3) discussed above: so long as some basic protection of

property rights exists, even formal contract enforcement may not be needed in a small

economy ­ self-enforcing relation-based governance can work well. But to go beyond

that stage, formal institutions of laws, regulatory agencies, courts, and the police are



                                               20

needed, and a malfunction in any part of this complex can create serious obstacles and

setbacks to growth.


        This argument can be linked to the idea of strategic substitutes and complements.

Where different institutions are substitutes, albeit imperfect ones, it is fine to start with

the simpler one and then move to a more complex and better one. But if the institutions

are mutually complementary, it is necessary to implement them in a package; for

example, privatization may fail unless there is an adequate structure of regulation, and a

good forum for adjudicating commercial disputes may be futile unless there exists a way,

public or private, to enforce the decisions of this forum. Therefore, to implement this

recommendation for gradualism successfully, a policy practitioner must identify which

policies or institutions are substitutes and which ones are complements. This is a difficult

task, for which the existing literature offers only a few and incomplete hints.


D. Imitation versus Innovation


        Linked to the conflicting arguments and evidence in the sections B and C above is

the question of whether to use the readily available models of Western legislative and

judicial institutions or to develop ones specifically tailored to ones country. Many of the

pertinent issues have been outlined above, so to save space I will not discuss them again

but merely mention their bearing on this question.


E. Crises ­ Good or Bad?


        Also linked to the question of speed and comprehensiveness of reforms is that of

the possibly beneficial consequences of crises. Olson (1982) presented this thesis in the

clearest form. According to him, many inefficiencies in prevailing institutions arise to

serve organized special interests, and the entrenched political power of these groups

prevents reforms that would be beneficial in the aggregate. Crises weaken or dissolve

these special interests, and allow a fresh start. Williamson (2000, p. 598) makes some

related observations. And Calvo (2005) points out that many recent crises have been



                                              21

followed by a sharp recovery in which the usual factors like domestic bank credit, current

account deficit, and investment were only small contributors ­ a miracle of a "Phoenix-

like" rise from ashes.

        It is hard to imagine a country deliberately risking a grave crisis in order to start

the Olsonian process; this might be called the "We had to destroy that country in order to

save it" recipe. Moreover, other researchers argue that crises are not necessarily

beneficial. The theoretical model of Hsieh (2000) clarifies the issues. The dynamics

negotiation on policy reforms following a crisis is a "war of attrition" or "chicken" game;

the question is who will concede first. With incomplete information about the

alternatives, value systems, and impatience of other players, such games can have

equilibrium strategies that involve lengthy periods without agreement, and may

sometimes result in total breakdowns. Whether crises change these equilibria for the

better is ambiguous. Dramatic practical examples of crises leading to total collapse are in

Diamond (2004).


F. Population ­ Resource or Curse?


        Thus far I have outlined several contrasting results and conflicting prescriptions in

the literature on institutions and their reform. A similar situation also exists with regard to

policies on specific problems. Population is a good example. A general presumption

dating back to Malthus is that a large and rapidly growing population is bad for economic

success. In the standard neoclassical growth model, a larger rate of population growth

leads to a steady state with lower output per capita. When one considers depletion of

natural resources, the consequences can be more drastic; Diamond (2004) is only the

latest of a long line of prominent scientists who have sounded alarms on this issue.


        But once again, there are strong counterarguments. Boserup (1981) has argued

that a high density of population is conducive to faster technical progress. And growth

models that incorporate such positive feedbacks, for example Lee (1988) and Kramer

(1993 b), find superexponential growth. Platteau (2000) finds that high density is also

conducive to the emergence of better networks and institutions.



                                              22

G. Whom to Believe, and What to Do


        I have sketched pairs of starkly conflicting research findings on a number of

important issues of institutions and policy. This can leave a user who is not an expert in a

particular area in a thorough state of confusion and indecision. Perhaps that is no bad

outcome. In any such situation, readers may find one side to be more convincing, but

rarely are the arguments and evidence from the other side clearly wrong or

unambiguously inferior, and many readers will differ in their judgment. Confronted with

such uncertainty about the true model and the factual details, practitioners who judge the

balance of considerations to favor one side would nonetheless be wise to proceed with

caution. The decisions being made here have immense consequences, and they are very

difficult to reverse. Therefore any uncertainty as to which is the better choice weighs

heavily in the decision. The favored choice must be not merely better than the other, but

better by a sufficiently large margin, to justify going ahead. Otherwise waiting for better

information is optimal; there is an option value to waiting. (See Dixit, 1992 or Dixit and

Pindyck, 1994.) This important lesson of the "real options" theory of irreversible

decisions under uncertainty is very relevant in this context.


        There is another equally important problem when it comes to action. Even if you

are convinced by one side in the argument, and want to follow it through into institutional

reform or policy action, there may not be much you or almost anyone can do. Suppose

you are the Minister for the Economy in an authoritarian regime. You read the writings of

Rodrik and others about the virtues of democracy, and are thrilled by the thought of

having this "meta-institution" that harnesses "local knowledge" in your country. What do

you do? Of course, if you are the Minister for the Army in a democracy and are

convinced by the pro-authoritarianism arguments, your may find it a little bit easier to

implement your favored institutional reform!


3. Shifts of Consensus




                                              23

        In the previous section, I laid out several stark disagreements that exist among

researchers at the same time, and showed how difficult it is to find reliable and feasible

policy recommendations from this chaos. This "cross-section problem" is augmented by a

"time-series problem." Among researchers as well as practitioners, there may be a

reasonable consensus at any one time, but the consensus can shift dramatically over time.

Here is a very brief outline of some such changes, and the warning they should convey.


A. Chasing the Latest Success Story


        At any time, some country is doing well, and academic as well as practical

observers are tempted to generalize from its choices and recommend the same to all

countries. After a decade or two, this country ceases to do so well, some other country

using some other policies starts to do well, and becomes the new star that all countries are

supposed to follow.


        In the 1950s, many intellectuals expected the U.S.S.R. to grow rapidly, and some

form of central planning, including import-substitution and state investment in heavy

industry, was offered as the key to success. Starting in the 1970s through the 1990s,

Japan and other emerging East Asian economies became the heroes, and markets and

openness became the vogue. Now China is the great example to follow. What next?


B. Role of International Institutions and Organizations


        For the last 50 years or longer, the World Bank, the IMF, and other international

institutions and organizations have played significant roles in giving advice, finance, and

aid to less-developed countries and transition economies. But there is shifting perception

as to exactly what their role should be. At one time they were supposed to be the

enforcers and guarantors of a country's commitment to policy reform; conditionality was

the key. But the conditions were criticized as "one-size-fits-all" recommendations.

Designing policies differently in each country based on local knowledge to fit its

conditions was argued to be vital, and countries' "ownership" of reforms became the



                                              24

buzzword. I like the idea that these institutions can act as facilitators or "catalysts"

(McCarthy, Bader, and Pleskovic, 2003); it can combine the features of discipline and

ownership. And it conjures up a nice image of a furnace or cauldron from which shining

reform will emerge. Dollar and Levin (2005) present evidence that the World Bank's

project finance and aid are most productive in countries with good institutional quality;

thus countries and international institutions appear to be strategic complements in the

production of economic success.


4. The Napoleon Prescription


         Faced with all these contradictions and shifts, I can identify only one consistently
valid policy prescription.4 It is the quality Napoleon valued most in his generals, namely

luck. Researchers want to identify causes, and practitioners want to know what they can

choose and change; therefore both sides may have neglected the important role that luck

has played in many countries' development successes or failures. Easterly (2001, chapter

10) is rare among economists in giving luck a substantial role and discussing it in

considerable detail. I offer a few supplementary examples.


         Japan's success with the auto industry owes much to luck. The improvement in

the quality of their small cars came at just the right time in the mid- to late-1970s, when

gas price rises had shifted the U.S. demand toward small cars and the U.S. industry had

not shifted its designs and production to match. And the Japanese quality upgrading in

the early 1980s also hit it just right, as the "voluntary" export restraints imposed by the

U.S. on Japan added an equal absolute shadow value of the quota to the prices of all cars

and thereby changed the relative price ratio in favor of larger cars. As for bad luck, think

of the many instances where large investments of resources and time in the building of

tourism in a country go to waste because of some natural disaster or political turbulence.




4 I have left out a few even more whimsical ideas, such as "The Mouse That Roared" recipe. In that movie,
a small country in Europe (the Dutchy of Grand Fenwick) declared war on the United States, with the idea
that when it lost the war, the United States would give it large amounts of aid. Unfortunately for the plan,
Grand Fenwick won.


                                                     25

5. Division of Labor and Diagnostics


         Why do we find so many contradictions in the research on the determinants of

economic growth and development, and corresponding conflicts in the recipes for

development success? Some of the explanation can be found in the diversity of theories

and perspectives that different economists and academics from other field bring to this

inquiry: development economics is undergoing a paradigm shift, from theories where

resource and technology constraints were the key obstacles to growth, to theories where

information asymmetries constrained the operation of markets, and now to theories where

institutions are the key. One must expect that in the process of a paradigm shift, no one

view will dominate and major differences will persist for a while. Of course this is not

very satisfactory for practitioners, who cannot afford to wait until the dust settles and a

new paradigm emerges.


         Another part of the explanation for the current state of differences can be found in

the diversity of data sources and periods over which econometric studies have been

carried out. But much of the explanation is surely to be found in the diversity of the

countries themselves. Each has some special aspects of its history, geography, religion,

society, polity, culture, and other features that influence how well or poorly a particular

prescription for institutional reform or policy change will fare when applied to it.


         Another way of saying this is that econometric and theoretical studies alike are

not the best way to generate policy prescriptions. Most of the cross-country regressions

fit far from perfectly; the myriad explanatory variables that have been tried explain only a

fraction of the variance. Theoretical modeling explores the implications of one cause or

mechanism in depth, deliberately isolating it from others, whereas policy prescriptions

require one to look at one country and consider all the different causes or mechanisms at

work there and how they interact with one another. The question policy prescribers must

address is not "What creates success on average across countries?" but "What is going

wrong in this country and how can we put it right?"




                                              26

          This seems to suggest a division of labor ­ academic researchers should look for

general tendencies and patterns, and should not offer specific prescriptions for individual

countries. That should be left to policy practitioners in the countries or in international

organizations who have much more detailed knowledge of the circumstances and needs

of individual countries. Of course some academics do combine theoretical knowledge

with deep knowledge of some countries; for example Fafchamps (2004) examines many

African countries through the lens of game-theoretic modeling of self-enforcing

governance mechanisms, and comes up with different prescriptions for three countries in

different circumstances and at different stages of institutional evolution.               5




          How can practitioners combine their knowledge with academic research? Must

this be a totally ad hoc matter, or can one develop something of a system for it? I would

like to suggest a framework building on the idea of "growth diagnostics" introduced by

Hausmann, Rodrik, and Velasco (2004). Their basic idea is to identify the crucial or

binding constraints in any situation, and focus attention on remedying those. They

develop a theoretical framework for this, and operationalize it using a "tree" much like

the branching decision diagrams one finds in the "troubleshooting" sections of product

manuals ­ "Does the computer produce four beeps when switched on? If yes, try such

and such next; if no, ..." Most importantly, Hausmann et al. ask the practitioner to look at

other symptoms that may be associated with low growth, and use a succession of these to

pin down a cause: "Are returns to investment low? If so, is the problem a micro risk like

corruption or a macro risk like financial instability? Or is it the case that investment

projects with high potential returns are available but are not undertaken because the cost

of finance is too high? And is the source of the high cost of finance to be found in local or

international conditions? ..."


          This approach has much to recommend it. Most importantly, it looks at multiple

dimensions of economic outcomes and tries to narrow down from a set of multiple


5Even such detailed and well thought-out combination of academic and field research can be blown off-
course by unforeseen shocks. For example, Fafchamps regards Zimbabwe as having relatively good
institutions and offers several suggestions for building on this basis to improve their financial
intermediation. Just a few years later, Zimbabwe's problems look quite different.


                                                      27

possible causes. But its sequential mode of thinking based on a tree structure seems

problematic. In reality, each case of development failure may have multiple causes acting

simultaneously. A related but more general point concerns the idea of "diagnosis" itself.

Dictionary definitions of the word are some variants of "identification of a disease or

disorder based on review of signs, symptoms, and laboratory findings." This sounds like

a problem of Bayesian inference, and is perhaps better captured in a table than in a tree.


         Table 1 shows such a schematic representation. The first column has the various

causes C1 , C2 , ... in its rows; each "cause" could in fact be a grouping or composite of

several causes. And the listing can be quite wide and comprehensive; it can include

political and institutional obstacles as well as specific policy errors of commission or

omission. The second column has the prior probabilities 1 , 2, ... with which the

various causes can occur. Historical analyses can give us estimates of these probabilities;
specific knowledge of a country can help us refine these estimates.6 The rest of the

columns have as their headings the various conceivable outcomes E1 , E2 , ... in the top

row; again each outcome could be a cluster or composite of several outcomes. The entries

in the cells are then the conditional probabilities, Pij being the probability that outcome Ej

will occur when cause Ci is present. In principle, econometric research can give us such

probabilities, although this is not how regression results are usually presented.


                          Table 1 ­ Causes, Effects, and Probabilities




   Causes         Prior probs.                                     Effects

                                           E1               E2                  ...               En

      C1                 1                 P11              P12                 ...              P1n

      C2                 2                 P21              P22                 ...              P2n

      ...               ...                ...              ...                 ...               ...

     Cm                 m                 Pm1               Pm2                 ...              Pmn



6 Medical diagnostics proceed using just such a combination of general statistical information about the
relevant population and the doctor's specific knowledge of an individual patient.


                                                    28

        If we observe a particular effect, say E7 , then the Bayesian posterior probability

that a particular cause, say C5 , is present becomes
                                             5 P5 ,7

                                             m

                                                iPi ,7
                                            i=1

If we want to be nearly certain whether or not a cause, say C5 , is present, we need to find

an outcome, say E7 , which will more typically be a cluster of outcomes or symptoms and

might be called a "syndrome" in a continuation of the medical analogy, such that

        (1) it is very unlikely to occur when the underlying cause is any other cause, that

is, Pi7 is close to zero when i is not equal to 5, so the posterior probability of C5 is close

to 1 conditional on observing E7 , and

        (2) it is very likely to occur when C5 is present, that is, P57 is close to 1, so the rest

of the P5j'sare close to zero, and if some other effect is observed, the posterior probability

of C5 becomes close to 0.


        The ultimate aim would be to find a complex of causes, or a syndrome, that

satisfies both these conditions. Then we can reduce the table to a two-by-two case, where

the syndrome we have identified is labeled C1 and everything else is, whereas the

outcome or effect we are trying to explain is E1 and everything else is E2 . In that ideal

setting, the conditional probabilities would be P11 = P22 = 1 and P12 = P21 = 0. Then the

posterior probabilities are 1 for C1 if E1 is observed and for C2 if E2 is observed, so the

outcome pins down the cause precisely. In reality we are unlikely to get a situation where

causes can be discerned from syndromes quite so well, but this gives us an ideal to work

toward.


        I hope this very rough conceptual scheme can be developed into something more

concrete and that it proves useful, not merely for thinking about the causes of some

development problems, but also for identifying and then tackling them in practice. If this

works, I will be delighted to have concluded a largely nihilistic tour of contradictions,

conflicts, and confusion in the literature on a positive note.




                                               29

References


Acemoglu, Daron. 2003. "Why Not a Political Coase Theorem? Social Conflict,

         Commitment and Politics." Journal of Comparative Economics 31 (4): 620-652.


--------- , Simon Johnson, and James A. Robinson. 2001. "The Colonial Origins of

         Comparative Development: An Empirical Investigation." American Economic

         Review, 91 (5): 1369-1401.


--------- , --------- , and --------- . 2002. "Reversal of Fortune: Geography and Institutions

         in the Making of the Modern World Income Distribution." Quarterly Journal of

         Economics, 117 (4): 1231-1294.


--------- , --------- , and --------- . 2005. "A Response to Albouy's `A Reexamination Based

         on Improved Settler Mortality Data' ". Cambridge, MA: MIT Working Paper.


Albouy, David. 2004. "The Colonial Origins of Comparative Development: A

         Reexamination Based on Improved Settler Mortality Data." Berkeley, CA:

         University of California, working paper.


slund, Anders. 1995. "The Case for Radical Reform." In Larry Diamond and Marc F.

         Plattner, eds., Economic Reform and Democracy, Baltimore, MD: Johns Hopkins

         University Press, 74-85.


Baker, George, Robert Gibbons, and Kevin J. Murphy. 1994. "Subjective Performance

         Measures in Optimal Incentive Contracts." Quarterly Journal of Economics 109

         (4): 1125-1156.


Bardhan, Pranab. 2005. Scarcity, Conflicts, and Cooperation: Essays in Political and

         Institutional Economics of Development. Cambridge, MA: MIT Press.




                                                  30

Barro, Robert J. 1999. Determinants of Economic Growth: A Cross-Country Empirical

         Study. Cambridge, MA: MIT Press.


Berglöf, Erik, and Stijn Claessens. 2004. "Corporate Governance and Enforcement."

         World Bank, Policy Research Working Paper No. 3409, September.


Bernstein, Lisa. 1992. "Opting Out of the Legal System: Extralegal Contractual Relations

         in the Diamond Industry." Journal of Legal Studies 21 (1): 115-157.


--------- . 2001. "Private Commercial Law in the Cotton Industry: Creating Cooperation

         through Rules, Norms, and Institutions." Michigan Law Review 99 (7): 1724-

         1790.


Besley, Timothy, and Robin Burgess. 2002. "The Political Economy of Government

         Responsiveness: Theory and Evidence from India." Quarterly Journal of

         Economics 117 (4): 1415-1451.


--------- , Torsten Persson, and Daniel Sturm. 2005. "Political Competition and Economic

         Performance: Theory and Evidence from the United States." London School of

         Economics, working paper.


Boserup, Esther. 1981. Population and Technology. Oxford, UK: Blackwell.


Boycko, Maxim, Andrei Shleifer, and Robert Vishny. 1995. Privatizing Russia.

         Cambridge, MA: MIT Press.


Desai, Padma. 1995. Beyond Shock Therapy. In Larry Diamond and Marc F. Plattner,

         eds., Economic Reform and Democracy. Baltimore, MD: Johns Hopkins

         University Press, 101-111.




                                            31

De Soto, Hernando. 2000. Mystery of Capital: Why Capitalism Triumphs in the West and

         Fails Everywhere Else. New York: Basic Books.


--------- . 2004. "Law and Property Outside the West: A Few New Ideas About Fighting

         Poverty." In Marc A. Miles, ed., The Road to Prosperity: The 21st Century

         Approach to Economic Development. Washington, DC: The Heritage Foundation,

         99-119.


Diamond, Jared. 1997. Guns, Germs, and Steel: Fates of Human Societies. New York:

         W. W. Norton.


--------- . 2004. Collapse: How Societies Choose to Fail or Succeed. New York: Viking.


Dixit, Avinash K. 1992. "Investment and Hysteresis." Journal of Economic Perspectives

         6 (1): 107-132.


--------- . 2004. Lawlessness and Economics. Princeton, NJ and Woodstock, UK:

         Princeton University Press.


--------- and Robert S. Pindyck. 1994. Investment Under Uncertainty. Princeton, NJ:

         Princeton University Press.


Dollar, David, and Victoria Levin. 2005. "Sowing and Reaping: Institutional Quality and

         Project Outcomes in Developing Countries." World Bank, Policy Research

         Working Paper No. 3524, February.


Easterly, William. 2001. The Elusive Quest for Growth. Cambridge, MA: MIT Press.


Engerman, Stanley L. and Kenneth L. Sokoloff. 1997. "Factor Endowments, Institutions,

         and Differential Paths of Growth Among New World Economies: A View from




                                            32

         Economic Historians of the United States." In Stephen Haber, ed., How Latin

         America Fell Behind. Stanford, CA: Stanford University Press.


--------- and --------- . 2002. "Factor Endowments, Inequality, and Paths of Development

         Among New World Economies." Economia 3 (Fall): 41-109.


--------- and --------- . 2003. "Institutional and Non-Institutional Explanations of

         Economic Differences." Cambridge, MA: NBER Working Paper No. 9989


--------- and --------- . 2005. "Colonialism, Inequality, and Long-Run Paths of

         Development." Cambridge, MA: NBER Working Paper No. 11057.


Fafchamps, Marcel. 2004. Market Institutions in Sub-Saharan Africa: Theory and

         Evidence. Cambridge, MA: MIT Press.


Gallup, John L., Jeffrey D. Sachs, and Andrew D. Mellinger. 1999. "Geography and

         Economic Development." International Regional Science Review 22(2): 179-232.


Giavazzi, Francesco, and Guido Tabellini. 2004. "Economic and Political

         Liberalizations." Milan: IGIER Working Paper No. 264.


Glaeser, Edward, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer.

         2004. "Do Institutions Cause Growth?" Journal of Economic Growth 9 (2): 271-

         303.


Greif, Avner. 1993. "Contract Enforceability and Economic Institutions in Early Trade:

         The Maghribi Traders' Coalition." American Economic Review 83 (2): 525-548.


--------- . 1994. "Cultural Beliefs and the Organization of Society: A Historical and

         Theoretical Reflection on Collectivist and Individualist Societies." Journal of

         Political Economy 102 (5): 912--950.



                                                33

--------- . 1997. "On the Interrelations and Economic Implications of Economic, Social,

         Political and Normative Factors: Reflections from Two Late Medieval Societies."

         In . John N. Drobak and John V. C. Nye, eds., The Frontiers Of The New

         Institutional Economics. San Diego, CA: Academic Press, 57-84.


--------- . 2004. "Impersonal Exchange without Impartial Law: The Community

         Responsibility System." Chicago Journal of International Law 5 (1): 109-138.


Hall, Robert E. and Charles I. Jones. 1999. "Why Do Some Countries Produce So Much

         More Output Than Others?" Quarterly Journal of Economics 114 (1): 83-116.


Hausman, Ricardo, Dani Rodrik, and Andrés Velasco. 2004. "Growth Diagnostics."

         Harvard University, Kennedy School of Government. Working paper.


Hay, Jonathan R. and Andrei Shleifer. 1998. "Private Enforcement of Public Laws: A

         Theory of Legal Reform." American Economic Review 88 (2): 398-403.


Heybey, Berta, and Peter Murrell. 1999. "The Relationship between Economic Growth

         and the Speed of Liberalization." Journal of Policy Reform 3(2): 121-137.


Hoff, Karla. 2003. "Paths of Institutional Development: A View from Economic

         History." World Bank Research Observer 18(2), 205-226.


--------- and Joseph E. Stiglitz. 2004. "After the Big Bang? Obstacles to the Emergence of

         the Rule of Law in Post-Communist Societies." American Economic Review 94

         (3): 753-763.


Hsieh, Chang-Tai. 2000. "Bargaining Over Reform." European Economic Review 44 (9):

         1659-76.




                                              34

Islam, Roumeen. 2003. "Do More Transparent Governments Govern Better?" World

         Bank, Policy Research Working Paper No. 3077, June.


--------- . 2004. "What Are the Right Institutions in a Globalizing World? And Can We

         Keep Them If We've Found Them?" World Bank, Policy Research Working

         Paper No. 3448, November.


Kaufman, Daniel, and Aart Kraay. 2002. "Growth Without Governance." Economia 3:

         169-229.


Keefer, Philip. 2004 a. "What Does Political Economy Tell Us About Economic

         Development ­ and Vice Versa?" Annual Review of Political Science 7: 247-272.


--------- . 2004 b. "A Review of the Political Economy of Governance: From Property

         Rights to Voice." World Bank, Policy Research Working Paper No. 3315, May.


Kremer, Michael. 1993 a. "The O-Ring Theory of Economic Development." Quarterly

         Journal of Economics 108 (3): 551-75.


--------- . 1993 b. "Population Growth and Technological Change: One Million B.C. to

         1990." Quarterly Journal of Economics 108 (3): 681-716.


La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny. 1998.

         "Law and Finance." Journal of Political Economy 106 (6): 1113-1155.


--------- , --------- , --------- , and --------- . 1999. "The Quality of Government." Journal of

         Law, Economics, and Organization 15(1): 222-279.


Lee, Ronald. 1988. "Population Growth and Induced Technical Change: Their Interaction

         in the Accelerating Stage." Mathematical Population Studies 1 (3): 443-465.




                                                     35

Li, John Shuhe. 2003. "The Benefits and Costs of Relation-Based Governance: An

         Explanation of the East Asian Miracle and Crisis." Review of International

         Economics 11(3): 651-673.


McCarthy, F. Desmond, William Bader, and Boris Pleskovic. 2003. "Creating

         Partnerships for Capacity Building in Developing Countries: The Experience of

         the World Bank." World Bank, Policy Research Working Paper No. 3099, July.


McMillan, John. 2003. Reinventing the Bazaar. New York: Norton.


Murrell, Peter. 1992. "Evolutionary and Radical Approaches to Reform." Economics of

         Planning 22(1): 79-95.


Olson, Mancur. 1982. The Rise and Decline of Nations. New Haven, CT: Yale University

         Press.


--------- . 1993. "Dictatorship, Democracy, and Development." American Political

         Science Review 87 (3): 567-576.


Persson, Torsten. 2005. "Forms of Democracy, Policy, and Economic Development."

         Cambridge, MA: NBER Working Paper No. 11171.


Platteau, Jean-Phillippe. 2000. Institutions, Social Norms, and Economic Development.

         Amsterdam: Harwood Academic Publishers.


Pritchett, Lant. 2003. "A Toy Collection, A Socialist Star, and a Democratic Dud?

         Growth Theory, Vietnam, and the Philippines." In Dani Rodrik, ed., In Search of

         Prosperity: Analytical Narratives on Economic Growth. Princeton, NJ: Princeton

         University Press.




                                              36

Qian, Yingyi. 2003. "How Reform Worked in China." In Dani Rodrik, ed., In Search of

         Prosperity: Analytical Narratives on Economic Growth. Princeton, NJ: Princeton

         University Press.


Rodrik, Dani. 1995. "Getting Interventions Right: How South Korea and Taiwan Grew

         Rich." Economic Policy 20: 53-97.


--------- . 2000. "Institutions for High-Quality Growth: What They Are and How to

         Acquire Them." Studies in Comparative International Development 35 (3): 3-31.


--------- . 2001. "Development Strategies for the 21st Century." In Boris Pleskovic and

         Nicholas Stern, eds., Annual World Bank Conference on Development Economics

         2000. Washington, DC: The World Bank.


--------- , ed. 2003. In Search of Prosperity: Analytical Narratives on Economic Growth.

         Princeton, NJ: Princeton University Press.


--------- . 2004 a. "Getting Institutions Right." Harvard University, Kennedy School,

         Working paper.


--------- . 2004 b. "A Practical Approach to Formulating Growth Strategies." Harvard

         University, Kennedy School, Working paper.


--------- . 2005. "Growth Strategies." In Philippe Aghion and Stephen Durlauf, eds.,

         Handbook of Economic Growth, North-Holland, forthcoming.


--------- and Romain Wacziarg. 2004. "Do Democratic Transitions Produce Bad

         Economic Outcomes?" Harvard University, Kennedy School, Working Paper.


Rubin, Paul H. 1994. "Growing a Legal System in the Post-Communist Economies."

         Cornell Journal of International Law 27 (1): 1-47.



                                              37

Sachs, Jeffrey D. and Andrew M. Warner. 2001. "The Curse of Natural Resources."

         European Economic Review 45 : 827-838.


Shleifer, Andrei and Daniel Treisman. 2000. Without a Map: Political Tactics and

         Economic Reform in Russia. Cambridge, MA: MIT Press.


Spiller, Pablo T., and Mariano Tommasi. 2003. "The Institutional Foundations of Public

         Policy: A Transactions Approach with Application to Argentina." Journal of Law,

         Economics, and Organization 19(2) Special Issue: 281-306.


Williamson, Oliver E. 2000. "The New Institutional Economics: Taking Stock, Looking

         Ahead." Journal of Economic Literature 38 (3): 595-613.


Widner, Jennifer. 2000. "Are Specialized Courts the Right Approach to Effective

         Adjudication of Commercial Disputes in Developing Countries?" In Gudrun

         Kockendorfer-Lucius and Boris Pleskovic, eds., The Institutional Foundations of

         a Market Economy. Berlin Workshop Series, Washington, DC: World Bank and

         InWent.


World Bank. 1996. World Development Report: From Plan to Market. New York:

         Oxford University Press.


--------- . 2002. World Development Report: Building Institutions for Markets. New York:

         Oxford University Press.




                                          38