A quarterly magazine of the IMF

June 2006, Volume 43, Number 2

Straight Talk

Crabs in a Bucket

Why constituencies are as important as constitutions in battling underdevelopment

Raghuram Rajan

Why do poor countries remain stubbornly underdeveloped? A growing consensus in recent years suggests the answer is that poor countries lack the necessary institutions or, worse, have the wrong institutions for economic growth. Indeed, recent empirical work suggests that better institutions seem to accompany both economic growth and greater stability. For example, a country with a responsible legislature and a sound tax system can respond to an adverse shock to revenues by raising taxes instead of by printing money. If the country has good institutions, it survives the shock by allocating the burden to those who can best bear it and in a way that least impinges on production; the country with bad institutions cannot pass responsible legislation and succumbs to hyperinflation, which not only reduces growth but also imposes the greatest burden on the poorest members of society.

That institutions matter cannot, by itself, explain underdevelopment. Development would then be simply a matter of setting up the right institutions. We need another explanation—perhaps that bad institutions are persistent. One reason may be that they are really hard to change; they are ingrained in the national psyche. Some former colonies have an easy culprit: their colonial masters and the institutions of exploitation they left behind. Alternatively, bad institutions may be self-preserving because they create their own support. For example, laws biased toward the elite will strengthen the power of the elite, who, in turn, will ensure that those laws aren't changed.

These two reasons for why bad institutions persist offer radically different views of development. The first offers little hope. We're the product of our history, and history can't be changed. It leads to a culture of blame where all evil lies in the colonial past and where current generations bear little personal responsibility. The second offers perhaps too much hope: change the institutions to unleash a virtuous cycle of growth and development. Change typically will have to come from outside, since internal structures are self-perpetuating. So, for example, the outside world must stop wars, impose democracy, provide aid, and fight corruption to set a country on the path to sustainable development. Given how bleak the first view is, it's no surprise that the second view predominates in the development community.

Yet doubts are growing about whether poor institutions are indeed fundamentally responsible for underdevelopment. Do institutions have a life of their own, or do they simply reflect a society's underlying power structures? Countries may be authoritarian, not because they don't have the right democratic institutions—that is, constitutions—but because the underlying structure of power groups—that is, constituencies—doesn't support democracy. Outsiders can impose democracy temporarily, but it won't catch on or won't create good outcomes unless the structure of society changes.

Perhaps the most ironic example of the inadequacy of the power of constitutions is Liberia, a state founded by freed American slaves. Its 1847 constitution was based on the U.S. ideals of popular sovereignty, separation of powers, and limited government. In some ways, it was more progressive in that it protected women's rights and abolished slavery. Unfortunately, Liberia itself was divided, with the descendants of the freed slaves forming an elite that lorded over the indigenous population. A League of Nations report in the 1930s found that Liberia represented "the paradox of being a Republic of 12,000 citizens with 1,000,000 subjects." It deplored the unsanitary conditions under which much of the tribal African population lived and termed Liberia's financial situation "tragic" and its monetary system "confusing." Despite some improvement in the middle of the 20th century and massive infusions of foreign aid, Liberia has not been a model state, to put it mildly.

I'm not saying that laws restraining the arbitrary powers of the government (or protecting citizens' property rights) don't matter for economic development, though the objective can often be achieved in ways other than simply through mimicry of the experience of the West. I'm saying laws will remain words on a piece of paper unless large and powerful constituencies in the population want them enforced. Unfortunately, in many poor countries, such constituencies don't exist.

Why the status quo may persist

The conundrum in development is why, especially after the advent of democracy, the large numbers of the exploited in poor but democratic countries don't combine to vote out the elite who exploit them. Why don't they change the system to provide opportunities for all? The easy but incorrect explanation, in my view, is that most democracies are sham democracies, where the exploited are easily fooled, have little money, and can't push for their own interests.

Instead, I believe that many poor countries are so riddled with inequality that no reform path commands obvious support and that the status quo persists despite being extremely inefficient. Let me offer a stylized example of what I mean. Consider a society with three "constituencies": a monopolist who owns all the factories, the educated middle class who occupy jobs as factory managers (and professional jobs, such as architects and doctors), and the uneducated poor who work in the factories. Suppose any two groups who vote for a reform can push it through (like all democracies, this is imperfect, with the rich having power because of their money and the poor having power because of their numbers).

Consider two reforms. First, promarket reforms allow anyone to open a factory in competition with the monopolist. Only the educated, however, can draw up the business plans and get the finance to take advantage of this opportunity. Second, education reforms allow everyone to get an education.

Clearly, the monopolist will oppose promarket reforms because he will face competition that will reduce his profits. And the educated will oppose education reforms because they will also experience competition (from the now-educated masses) for the lucrative jobs they currently occupy. But will either one get support to vote down the reforms they dislike?

The answer could well be yes. The monopolist would prefer to educate the poor for that would give him a larger labor pool to pick managers from, thus reducing salaries he has to pay. However, the monopolist also knows that if he votes to expand education, he will have a workforce (the formerly uneducated and the formerly educated) that is united in interests. This enlarged constituency will then push for promarket reforms. To forestall the greater loss from promarket reforms, the monopolist will align himself with the educated against expanding education.

If education reforms are unlikely to be enacted, the uneducated may reject promarket reforms, preferring the status quo instead. Although promarket reforms expand opportunities for the educated, they also have a dark side for the poor. Given that the educated have greater business opportunities, those among them who choose to continue providing services such as health care can demand higher fees. The uneducated, whose job opportunities go up only a little, if at all, may face a substantially higher cost of living because of the opportunities the educated now have. They may side with the monopolist in voting against promarket reforms.

Fix the constituencies

In sum, even in a society where political institutions ensure that citizens' preferences matter, initial inequalities (in education and wealth) may be self-perpetuating. Citizens, fearing that the advantage gained by one group may come at the expense of the meager rents of the other, become like crabs in a bucket, preventing each other from getting out. Uncertainty about who will get the benefits of reforms can further compound resistance. Underdevelopment can persist with the full connivance of the exploited, even with reasonably well-functioning political institutions. Finally, while stylized, the example is consistent with the evidence that far too many poor economies, like India, have underemphasized universal education while overemphasizing higher education and that the poor and uneducated in a number of countries in Latin America have turned against (partial) economic liberalization because they see few of the new opportunities while bearing additional costs.

What lessons does this suggest for development? Clearly, the answer "fix the political institutions" is probably incomplete, if at all correct. "Fix the constituencies" is probably more on the mark, but how? A number of development successes, like Korea, undertook serious land and education reforms prior to their takeoff, as have a number of the fast-growing Indian states. It seems that reforms reducing inequalities in factor endowments (like land) and those improving access to education and finance can strengthen the constituencies for broader economic liberalization. That is, the free-access economy may be a necessary stepping-stone to the free-enterprise economy. I should not, however, minimize the difficulty of enacting such endowment-spreading reforms in highly unequal societies. The bottom line is that development is likely to be a complex political process in which the people themselves must do much of the heavy lifting. The outside world can help at the margin but only if the people have ownership. And ownership—even of something as beneficial as development reforms—can't be taken for granted.