A caution against faddist approaches to industrial policy (the old industrial policy rules still apply)

It’s a good thing that World Bank is finally OK with industrial policy, but policymakers shouldn’t let themselves be constrained by the coming prescriptive “best practices.”

An Africanist Perspective is back after an unscheduled hiatus (life happens). Thank you for being a regular reader. If you haven’t done so yet, please hit subscribe to receive timely updates along with over 34,000 other subscribers.

I: Has industrial policy become the new development fad?

The March 31, 2026 release of a World Bank policy research report sympathetic towards industrial policy caused quite the buzz in development circles. Such a report has been a long way coming since the infamous 1993 misreading of the core drivers of East Asian economic takeoff; as well as countless efforts to stifle ambitious context-based policymaking before that. As most commentators on the report have already noted, the intellectual and empirical justifications for industrial policy are decades old, and includes lots academic/policy giants that were previously ignored by the Bank and the broader international development community. Better late than never, I guess.

This is post is not about the report itself (it’s a good primer on industrial policy for those unfamiliar with the concept), but mostly about what the World Bank (as well as donors and other multilaterals) might do with the recommendations in the report (and how policymakers in low-income countries should respond).

So why now? A cynical read of the situation might be that this is yet another instance of the World Bank as an institution (as opposed to the actual authors of the report) externalizing the dominant intellectual/policy orthodoxy among its most powerful shareholders. Industrial policy was bad when low-income countries were dabbling in it to accelerate their processes of catch-up development; but it’s now fine because certain high-income countries have reasons to do it more aggressively and openly (they’ve always done it, anyway) to compete with China.

Whatever the actual motivations behind the newfound institutional accommodation at the World Bank, the reality is that the last decade (especially with regard to green techno-industrial transformation and Artificial Intelligence) has created an intellectual cum political permission structure for anyone who wants to support industrial policy without fear of being considered “unorthodox” or a heretic. Industrial policy is now cool not only because versions of it make sense in both the real world and seminar rooms, but also because it makes perfect geoeconomic sense.

How should policymakers in low-income countries approach the openness to industrial policy at the Bank? I would suggest that they heed the lesson in this excerpt from Joe Studwell’s How Africa Works, in which he describes the faddist nature of the World Bank’s policy evangelism:

The aid sector […] is a fashion business. On a visit to the World Bank’s annual Land and Poverty conference in Washington D.C. at the start of my research for this book in 2019 I was told that just two things are required to make a poor country rich: land titling systems employing technology like drones and projects that empower women. Those causes owned the catwalk and filled the halls. Periodically, the dominant aid show changes.

‘The fashion when I joined [the World Bank] was community-driven development,’ Simeon Ehui, a regional director for Africa, told me. In part, fashion in the aid industry is driven by a natural human desire to identify a developmental magic bullet, even though none exists. In part it reflects the fact that donors are more responsive to simple, tidy solutions than to the reality that development is complex, requiring what the economist Albert Hirschman called ‘a multidimensional conspiracy’.

In aid, fashion is inevitable. Successful developing country governments are the ones able to decide what they need from the catwalk and politely decline what they do not.

More deleterious than fashion is ideology in aid. Ideologically driven aid offers only one option. Its heyday was the 1980s and 1990s, when the IMF and the World Bank pedalled a single flavour of financial sector reform that was born out of free market policies in Reaganite America and Thatcherite Britain. Developing countries were suddenly deemed to require the same mix of freely tradable currencies, private banks and lightly regulated stock markets that rich countries required.

For capable developing country governments nurturing infant export industries, it was terrible advice because it curtailed their capacity to finance and run industrial policy. Successful developing states ignored the advice. Less capable leaders swallowed the pill.

The point here is that any advice on industrial policy from the World Bank should be taken with a huge grain of salt.

The Bank may has some strengths on narrow technical questions, but it finds itself at sea on issues that require deep knowledge of context, politics, and history (like industrial policy). It’s for this reason that, as described in the above excerpt, policy prescriptions from the institution change with the seasons, with little or no consequences for past errors of commission or omission, however obvious those errors were ex ante.

The Ajaokuta steel plant in Nigeria has never produced steel despite gobbling up billions of dollars in public investments over decades. Source: Premium Times.

Implementing industrial policy is very hard — as demonstrated by examples like Ajaokuta. As should be obvious, countries that choose to do so should fully own and implement industrial policy with eyes wide open. In brief, this means having coalitions of elites who actually want to develop their societies, and not just a bunch of two bit rent-seekers. Successful industrial policy shouldn’t mean a suspension of the rules of economics (Africa is littered with too many examples like the infamous Ajaokuta steel plant already). Neither should it mean rushing to mimic whatever other countries are doing, or implementing “global best practices” even if (or especially if?) sanctioned by World Bank economists. Instead, it’s a guided management of parts of the process of development that should start with a full internalization of the rules of economics, and then going about tweaking them for context on the road to ever deeper market efficiency. Stated differently, the bending of certain rules of economics must always be justified by concrete and defensible developmentalist imperatives.

Importantly, countries must embrace the idea of learning by doing. Relying on off-the-shelf industrial policies almost guarantees failure. There’s no set recipe. All anyone knows are the ingredients.

All this to say that industrial policy as will likely be packaged and peddled by the World Bank will have strong faddist elements, and will therefore not be of much help to low-income countries. The next section explains this claim in more detail.

II: You can’t successfully implement industrial policy by adhering “international best practices”

Now that there’s intellectual tolerance for industrial policy, it’s very likely that we’ll see the emergence of industrial policy “expertise” at the World Bank and other multilaterals. This would be a waste of human and financial capital. There are two reasons not to institutionalize “industrial policy advisories” within multilateral institutions.

First, low-income countries simply cannot and should not outsource the ambition and implementation discipline required to successfully execute industrial policies. To reiterate, the best way to get good at it is through learning by doing (something the World Bank report acknowledges in Box 3.2). Among other things, this is because industrial policies are inherently political, and must be explicitly anchored in domestic political economies. You need a durable developmentalist political coalition, bringing together bureaucrats, businesspeople, the wider society, and politicians. You need to make policy choices that set you apart from peers (and not procrustean recommendations from Washington, DC or wherever). You need to be open to experimentation and making mistakes (extreme risk aversion in the name of chasing “optimal” policies is terribly overrated). And you need to be able to think and act with the long term in mind (as opposed to short project cycles). All this necessarily involves building the requisite competencies as you go — which means that you shouldn’t impose too many preconditions on the process.

In short, you can’t projectize industrial policy.

Second, organizations like the World Bank and most other multilaterals are not institutionally designed to support industrial policy. The project cycles — from design, to institutional approval processes (influenced as much by geopolitics as narrow/faddist concerns), to implementation — are way too rigid to work alongside the domestic political economy and bureaucratic demands of industrial policymaking. Layered on top of this is the fact that the Bank, for example, pretends to not engage in their clients’ domestic politics — yet getting the politics right is foundational to being able to execute successful industrial policy.

In addition, and let’s be blunt here, the Bank just doesn’t have the human capital to execute on industrial policy projects. Most economists at the institution were trained to be skeptical of or downright hostile to industrial policy (see, for example, the World Bank’s recent ill-timed commentary on Nigeria’s energy policy). Furthermore, much-needed capacity for contextual heavy lifting just doesn’t exist at the Bank in sufficient qualities or quantities. There are virtually no career incentives to invest in knowing particular countries and their political economy dynamics well enough to be of use over years of executing on an industrial policy strategy.

Consequently, the go-to implementation strategy would likely be to projectize industrial policy to death under reams and reams of “best practice” technical policy prescriptions that can quickly scale globally.

Finally, historical institutional disdain for contextual knowledge means that, in all likelihood, the Bank’s recommended industrial policy strategies will not be informed by an objective understanding of the causes of economic underperformance in low-income countries. Reading the report, it’s hard to not conclude that the Bank still espouses a fundamental misreading of low-income countries’ economic histories over the last 60 years.

Consider this bit from the report (the answer to “why now” from the report’s perspective):

Three broad shifts in recent decades warrant a reexamination of the skepticism surrounding industrial policy. First, the talent available to governments has expanded substantially amid rising education levels globally. Second, the political environment in many countries has become more supportive of development objectives. While attribution is difficult, broad improvements in macroeconomic stability, growth, and health outcomes likely reflect a closer alignment of politics with development goals, rather than improvements in talent alone. As a result, many political systems today appear more capable of implementing industrial policies effectively and efficiently. Third, most economies are now open rather than closed. This openness narrows the scope of industrial policy—placing greater emphasis on public inputs—but also reduces the need for any single agency to control every policy lever.

All three claims call for a rebuttal. First, levels of education attainment among policymakers/politicians weren’t the binding constraint to successful industrial policies over the last several decades. For the most part, ideas, political instability (i.e., absence of durable developmentalist coalitions), external shocks, and over reliance on multilateral policy advice ranked far higher. In fact, there are cases, like Kenya, where the quality and ambition of policymaking arguably peaked in the early 1970s.

Second, it’s unclear what is meant by “a closer alignment of politics with development goals.” Is this a nod to post 1990s electoralism as a driver of successful policymaking? If so, this claim is also historically inaccurate. For example, most African countries’ education enrollment rates (and related ambitions) first peaked in the late 1970s at the tail end of postcolonial big picture developmentalism (when many were autocracies), before plummeting during the dreadful long decade of underdevelopment (1980-1995). The recent enrollment uptick under MDG/SDG inspired UPE policies were merely a return to the late 1970s trend. Political demand for development has always existed in low-income countries. The problem has been on the supply side (and global shocks). More importantly, when thinking about industrial policy, what deserves emphasis is not electoralism or regime types, but the degree and quality of government (i.e., state capacity). On this score, it is true that things have improved significantly over the last few decades.

Third, the choice of observable metric of success — the performance of export-oriented firms — unnecessarily compressed what it takes to succeed at industrial policymaking (in this quote it’s couched in the language of trade trade openness, but in the executive summary it’s more explicitly stated as export-oriented firms’ performance). Strictly speaking, it is true that export orientation is the best way to discipline firms that receive policy support and force them to invest in productivity improvements. However, there are multiple steps between choosing to execute export-oriented industrial policies and having successful global firms — from labor policies, to allocation of capital, to choosing, supporting, and disciplining national champions, to learning and technology transfer, to the development of national/regional value chains etc. These process may take more than a decade, and may at first yield uneven results across the multiple pieces of the puzzle — an interval and degree of variation in policy outcomes that are well outside the World Bank’s institutional/project attention span.

Instead of merely asserting that things are different now (which is true!) and then listing a slew of potential policy options, it would’ve been helpful if the report spent time on elements of state capacity (especially the ability to at once help specific firms/sectors and also discipline them) and political coalition building for successful industrial policymaking (including how this might be shaped by the presence of land thereof of heightened electoral competition). Everything else flows from these two factors in a highly contextualized fashion; and without a handle on both there’s very little analytical mileage we can get from abstract discussions of institutions, embedded autonomy, incentives, and accountability.

Overall, the report treats politics as sand in the gears, and emphasizes the need to insulate implementing agencies from politics (see Chapter 5 on getting institutions right). While this is true, it’s worth noting that the insulation of public agencies doesn’t just miraculously happen. Political developmentalist coalitions must agree to such an arrangement. In other worlds, politics is a core feature of industrial policymaking, not a bug. And the fact that the Bank is likely to default to the “apolitical” approach to industrial policy advise is good reason to be skeptical of any such programmatic endeavor from the institution.

III: Conclusion

In general, I think it’s a good thing that we now have an ideational accommodation of industrial policy as part of the developmentalist policy toolkit — but perhaps for different reasons than many would assume. Above all, I see this opening as a chance for African policymakers to embrace ambition, focus on their objective developmental needs, and adopt aggressively pro-growth policies that are contextually suitable. The context-dependent logics of industrial policy demand that African policymakers eschew policy extraversion. Now is not the time to dilute their development ambitions through decidedly subpar remote control industrial policy. Again, you can’t projectize industrial policy.

Lastly, and I cannot stress this enough, I hope that African policymakers don’t view the recent ideational shift on industrial policy as a license to gratuitously expand public sector participation in their economies at the expense of the private sector. Obviously, the African state still has a huge role to play in the region’s development journey. Absolutely no doubt about that. However, the region’s states must also support the private sector. The journey to mass job creation and broad-based development in the region will necessarily have to pass through the emergence of large private sector firms. It’s as simple as that.