The recent World Economic Outlook report by the International Monetary Fund (IMF) has an interesting surprise. It includes a chapter on institutional reform that places responsibility for economic success or failure squarely on the side of institutions.

Institutions are defined as “the set of formal rules-and informal conventions-that provide the framework for human interactions and shape the incentives of members of society”. Not bad. Someone at the IMF has been reading Nobel Laureate Douglass North.

Good institutions, the report continues, are those that create “rent-free environments in which small groups are not able to take advantage of—for example—a monopoly position in a particular industry or activity, or privileged access to natural resources”. Countries that prosper are those in which the government neither bends the rules to favor certain groups nor limits the entry of those without power into the marketplace. If institutions in Africa could be improved to the level of developing Asia, Africa’s GDP would double, we are told.

So far, so good. The IMF looked at sixty-five countries that have undergone institutional reform in the last three decades, asking itself what determines institutional change. This is where the study runs into some trouble. To answer this question they conducted econometric exercises mixing the data from the various countries and coming up with certain patterns. The problem is that econometric exercises don’t really work with factors that have to do with ideas, choices, and historical contexts rather than numbers.

The IMF concludes that trade openness, freedom of the press, neighbors with good institutions, higher levels of education, and higher initial per capita income tend to determine institutional reform.

But a look at countries where institutional change has taken place indicates that most of these factors followed rather than preceded reform.

Opening countries to free trade is an institutional change in and of itself. Chile privatized companies and liberalized internal markets even before it opened the country to substantial international trade (only recently did they bring their tariffs down to an average of just over 2 percent). Estonia got rid of its tariffs in the initial stage of reform, starting a sequence of changes that subsequently encompassed other institutional areas.

Freedom of the press is extremely important, but it is not a precondition for institutional reform. There is no freedom of the press in China and that country’s dictatorial bureaucracy has been implementing reform since 1978.

Good neighbors are also not necessarily a great incentive. If that were the case, Mexico would have copied U.S. institutions a long time ago. Ireland was stuck next to Britain for centuries before it decided to catch up institutionally.

Higher levels of education are no precondition either. Spain had one of the lowest education levels in Western Europe when Franco started to open the economy and that had not changed by the time Felipe González got rid of many old institutions in the 1980s. The two most successful reformers in Africa-Botswana and Ghana-do not have higher education levels than dismal Latin American reformers like Honduras and Paraguay.

Finally, a higher initial per capita income can actually work against reform. Argentina was prosperous for half a century (including the early 20th century) and then produced Juan Domingo Peron, who destroyed that prosperity. Bolivia was the first Latin American democracy to engage in reform in the mid-1980s and it had the lowest per capita income in South America (the reforms were ultimately insufficient but so were those of other countries).

Here are two factors the IMF might consider as determinants of institutional reform in the future: leadership and crises. Most successful reformers had enlightened leadership in times of political, economic or social upheaval. The combination of the two factors created the conditions for reform. That is true historically as well as in contemporary times.

In the nineteenth century, the emergence of the Meiji leadership in Japan during a time of conflict with the U.S. after a long period of isolation created the conditions for modernization. In nineteenth century Britain, the relentless leadership of Richard Cobden and John Bright combined with the Irish potato famine forced the repeal of the Corn Laws.

In contemporary times, the leadership that brewed in opposition circles under Communism in Central Europe was able to take charge with the collapse of totalitarianism. I am thinking of Charter 77 in Czechoslovakia and the Budapest intellectuals in Hungary—as well as the Solidarity movement, a less intellectual but equally important leadership (initially, at least) in Poland. In New Zealand, the economic crisis of the 1980s combined with the emergence of a visionary leadership in the Labor Party (an unlikely catalyst of change) that included people like Roger Douglas gave rise to a free-market transformation. Even in China, the crisis brought about by the disaster of the Cultural Revolution and the emergence of Deng Xiao Ping, a very intuitive despot, generated reform.

In Latin America, the crisis of hyperinflation in the 1980s produced a struggle between statists who wanted to nationalize everything and protectionists who wanted to maintain private property. The result was preemptive reform led by the latter. It did not generate the desired results, but that is a different story.

Leadership and crises cannot be measured econometrically. But if we want to make reform more credible in the eyes of the skeptical, let’s be clear about its causes.