How to Beat the Odds: Reflections on Stefan Dercon’s "Gambling on Development"

In Stefan Dercon’s new book, Gambling on Development, he distils the lessons from some 30 years of research into global development and a decade as a senior technocrat in the UK’s development architecture (first Chief Economist in DFID and then Development Policy Advisor to the Foreign Secretary in FCDO) into a single thesis for how development happens.

His thesis is simple: development happens most reliably when local elites—the politicians, senior bureaucrats, power brokers, and sometimes military leaders who hold the reins of the country—want it to happen. These are the people who wield political and economic power, and when they decide their best bet for perpetuating their power and their own wellbeing is to pursue widespread gains that accrue to broad swathes of the population, they make good development outcomes more likely. But this is a gamble: by making the long-term bet on development, they are by definition not betting on, for example, the quick and easy smash-and-grab of kleptocracy, or the narrower returns to exploiting natural resources with just enough trickle-down and redistribution to stay in power. And their pursuit of development usually requires that each player in the elite coordinates on this equilibrium—any who defect and cause conflict can bring the whole system down. Even if they reject temptation and manage to coordinate, development is not guaranteed: it requires good fortune and good management. Nevertheless: when the elite bargain is not developmental (but instead pursues, for example, the self-enrichment of the elites through theft or patronage), development will not happen; and when it is, development becomes more likely and easier to support. He supports his thesis with a number of examples and counter-examples; stories, data, and political analysis.

I am not writing a review of Stefan’s book. He has been, at various times, my co-author, boss, mentor, research supervisor, and friend. I will simply recommend it strongly: I have read it in various versions at least three times, and I have seen the underlying thesis evolve over a decade of working together. Yet it still has the capacity to surprise me with its implications. But not all of these implications are straightforward; some help us understand the process of development but make murkier the process of supporting development. Others throw into sharp doubt the model of development that underpins how most donors operate. I discuss four implications of the book for how we think and do development (the CGD mission, after all) below.

Watch closely and move fast

In Gambling on Development, Dercon argues that elite bargains are rarely negotiated and announced in public, with the pros and cons for all players clearly spelled out. They must instead be inferred, even for those close to the key players. There is never a sheet of paper that tells you what the real objectives are, and what the real payoffs for different actors will be.

That means that when a development bargain is struck, it may—initially—be very difficult to recognize, or to understand how it is structured. That matters, because these development bargains are rarely perfect, as Dercon repeatedly emphasizes, which means that only some parts will merit a great deal of support from outsiders. At the same time the elements of the bargain that are developmental may not be immediately obvious, shrouded in less savoury aspects of the deal struck among the elites. Dercon gives the example of Indonesia, where a development bargain was struck in the 1970s, prompting sustained growth and poverty reduction, while at the same time providing opportunities for the elites to make an illicit killing, with Madame Tien Suharto becoming widely known as ‘Madam Ten Percent’ for her cut of the large state contracts being awarded. An outsider may not have recognized that both aspects were central to the bargain struck. Despite this difficulty, supporting the developmental part of the elite bargain may be important, since such bargains are likely to be fragile in their early stages, with their main payoffs tend to be long term rather than immediate.

Because of this, Dercon argues that donors should be ready and willing to put financial and technical resources behind the elements of the elite bargain that are developmental, to help sustain them. This puts the onus on them to keep a close eye on how the elite bargain evolves—not just through electoral politics, but in how the most powerful actors in the private sector, among government bureaucrats and the military engage with each other. But this suggests two problems: how do we identify a nascent elite bargain? And how do donors keep tabs on the workings of the elite? The book doesn’t have a neat answer for either, and perhaps this is because there isn’t one. But I would argue (as Dercon does) that donors should invest in this sort of local knowledge, through staffing, longer postings, and simply incentivizing staff to engage carefully with local actors.

Outsiders need to gamble too

One of the more compelling aspects of the development bargains theory is that they are gambles for domestic actors: there is no guarantee of rewards or success. But one of the implications of this is that outsiders who want to support a development bargain are also taking a gamble. In some ways it’s a double gamble: since they are faced with incomplete and asymmetric information about the real content of the elite bargain, donors are uncertain about the true nature of the elite bargain. They must first take a bet on whether a development bargain exists at all. And then, if they correctly diagnose the existence of a such a bargain, they need to gamble again by putting money behind it to accelerate progress, even into activities (such as policy advice and attempts to change and improve economic management) with uncertain prospects of success.

This double gamble is a very different approach to the general trend in how donors operate. Increasingly, donors look for cost-effective interventions for which data and studies from a range of contexts demonstrate good returns, as exemplified by the World Bank’s and FCDO’s recent ‘Smart Buys’ work in education. But the implication of Dercon’s thesis is that many kinds of support that fail in other contexts can work when a development bargain is in place. Support to economic reform may be good money after bad where kleptocracy and a terrible business environment is the game of the elite, aimed to raise the barriers to entry and maximise short term pay-offs to those on the inside, but may have a multiplicative effect on development and growth when a development bargain is in place. This suggests that at least part of donor portfolios should be devoted not to things that work broadly in a range of contexts, but to things that only work when a development bargain exists, if you think one is in the ascendancy. If the elites in a country are taking a long term bet on development, donors should be willing to match them, even at the risk of the bargain collapsing and achieving little.

Bet on bureaucrats

Gambling on Development is littered with stories about brilliant bureaucrats who manage to navigate complex elite bargains in such a way as to push for genuinely good economic and development outcomes, even if this is just to keep disaster just at bay. He talks about Riad Salamé, known as the Magician of Lebanon for his ability to (until 2019 at least) steer the country clear of economic disaster from the Central Bank; Ato Newai Gebre-Ab in Ethiopia, Meles Zenawi’s economic advisor; at his book launch in Oxford, he spoke admiringly of the late Benno Ndulu, former governor of the Bank of Tanzania. All of these policymakers combined technical brilliance with the ability to navigate their political environment; they understood the deals in place, where progress could be pushed, and how to push it.

There is another book to be written about how to be a great policy adviser in messy circumstances, one with relevance to literally every country in the world. But the implications for development actors are more immediate. They should bet on the right people as well as betting on good systems building. Of course, in the long run, a system that doesn’t depend on specific individuals to function well is preferred; but the long run can be a long time indeed. In the interim, finding the bureaucrats and teams that work and that have the ear of those in power but want to push for sensible economic policies that support growth and development is valuable, even if this means making compromises on the way. This has some kinship with the idea that to help programmes bed down and succeed, a local champion is needed; but it goes beyond this too, since this backing should be independent on any specific set of policies or programmes the donor is pushing.

Think carefully about support for the places left behind

One question the book never fully answers is how best to respond in places where no development bargain is in place, but no humanitarian crisis is unfolding; the places that are stable, poor, and in stasis. This is a particularly tricky question. One argument is that any support, including that directly delivered to the poor simply provides an excuse for the elite bargain to ignore them, or even worse directly props up anti-developmental elites. Even worse, does it help sustain non-developmental bargains? Does this mean that donors should be more sparing in their support?

It's not clear what the correct approach is. If it’s genuinely the case that a development bargain appears out of reach, or extremely unlikely, is it better to try and increase the chances of one happening, even from a low base, or to support those left behind by its absence? Thanks to cash transfers and the possibility of delivering good results through nongovernmental organizations and other non-state actors it is possible to achieve meaningful improvements in standards of living and key health, nutrition, and education outcomes despite an elite that doesn’t support them. Perhaps this isn’t a path to prosperity, but it may be a path out of destitution.

And there is a question of where the marginal dollar should go. If a development bargain is genuinely struck, and good technocrats are in place, is it grant aid that should be provided? Perhaps the role of donors is to help these places access market finance, while focusing their most concessional resources on people where the state gives them least help. This though, raises the spectre of a moral hazard: letting the poor languish in the knowledge that donors will step in.

Ultimately, the prescription is to take steps to bring a development bargain closer—Dercon suggests backing with those bureaucrats who pursue good economic outcomes within the existing elite bargain and working with firms directly to nudge the economy to a place where for the elites, placing a long-term bet on development starts to look more sensible and less like a wild gamble. Without one, development—and certainly prosperity—will be slow, painful, and probably out of reach.

This blog benefited from excellent comments from David Evans and Masood Ahmed. Any remaining errors are the author’s.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.