INDUSTRIAL POLICY IS TRICKY

An Everyday Guide to Development - Part I

Development is an important subject - and not just as an academic or curiosity exercise. The fate of billions of people depends on the ideas people come up with and experiment with in the policy space. I want to write a few quick (and hopefully handy) posts of the few things I have learnt on the subject and how I think lay people should evaluate the ideas floating around in this area. Note that this is not an exhaustive list of all there is to know on the subject. Development is hard, and societies are very complex. This is just a collection of personal observations to help “ginger” your consciousness.

INDUSTRIAL POLICY

Since Industrial Policy is all the rage now, with calls to ban one thing or another - it seems like the obvious place to start. Most Nigerians are Industrial Policy (sometimes IP for short) advocates and/or protectionists in all but name. Many of us cannot even comprehend the economic gains in its full complexity. But we mostly want local companies to make the things that we use and to also employ our relatives. There are two common ways of defending this view. One is that we think if the country has a particular input in abundance (e.g bamboo forests), then it makes no sense to import another commodity (e.g toothpicks) made from that input.

This is a hard argument to refute because it is intuitively appealing. Despite its early formulations dating over two hundred years, the theory of comparative advantage that refutes this view gets quite difficult to explain. But the second common argument advanced for industrial policy is more influential and has become unquestioned doctrine in some policy quarters. The assertion is that developing countries that achieved relative prosperity in the last fifty years, did so by engaging in industrial policy.

I want to focus on this second argument for it is more tractable and falsifiable - meaning that we can easily engage with the literature and evidence on whether industrial policy is a standard component in the toolbox of policymakers in the countries that are richer and more prosperous than others today. There is a large body of work on industrial policy in the sub-discipline of development economics, hence I find it easier to simply pick off the things I have learnt from my outside observation.

I have been going on about the phrase “industrial policy”, but what would be an understandable yet accurate description of the term? Industrial Policy is quite different from technocratic policy-making - like what goes on in a central bank. Industrial Policy is characterized by deliberate political intervention in the structure of an economy. This intervention may or may not be informed by evidence or general economic information. And it usually comes in form of various subsidy packages (intervention funds, tariff waivers, tax credits etc) to targeted sectors and industries.

The chart above is one of my favourite graphs in the world. It depicts one of the modern economic miracles in recent history. How did a country like South Korea manage to leave Nigeria far behind? There are many answers, and they are complicated. But there is some consensus that South Korea’s period of aggressive Industrial Policy is a critical factor in its prosperity. It is then very straightforward to conclude that Nigeria too can replicate some of this income growth by engaging in equally aggressive Industrial Policy. Simple as this conclusion might be, things get fuzzy when you dig into the details.

Evidence of the effects of IP is mixed, and causation can be notoriously hard to tease out. But there are some salient points from the literature that can act as a smell test when policymakers are trying to sell you.

1. EXPORT DISCIPLINE

In his book “How Asia Works”, Joe Studwell detailed how the industries that benefited from government subsidy were strictly pruned for exports. The underlying logic is that subsidizing domestic consumption does not represent good returns and can create a huge fiscal burden. Also, making things for the global market is the best way for domestic firms to build needed technical capacity that increase their productivity.

2. EXPORT CONTENT  

If we expand on the logic of export discipline, we can conclude that IP (when it works) is a good way for an economy to earn income that can be reinvested in public goods that will further rotate a virtuous cycle. What subsidized industries earn from foreign markets can be repaid to the government in interests and taxes, which can then be invested in infrastructure for other sectors of the economy. The foregoing reasoning then suggests that not only is it better make for exports, you should also make that which can earn you the most profit. Economists Dani Rodrik and Ricardo Hausmann concluded from their study that countries that benefited from IP are not those subsidizing agricultural production, rather they earn most of their dollars from making “high tech” products. These can range from semiconductor chips to automobile manufacturing or steel making and ship building. This does not imply that a country cannot compete for exports in other industries. It just means if a government will intervene and invest directly or subsidize, then it works in industries with the highest returns for capital and capabilities.

3. COMPETITION

There is a general wisdom in development circles that a government cannot (should not) successfully pick winners in an economy - especially with IP. Putting aside the corrupt motives of governments that picks winners. The question remains whether it’s the best way to do IP. Looking again at the East Asian example, available evidence points to competition as the best policy tool. Rather than protect an industry from import competition, it is not enough to make the recipient firms export facing - they also have to compete against each other. This is how they develop the needed capabilities to be global competitors. Global industrial giants like Samsung and Hyundai went through this process of intense local competition.

IP seems like a free lunch for advocates when the conditions listed above are met. But the policy world is never that simple. Development economist William Easterly once joked that “governments may not know how to pick winners, but winners know how to pick governments”. This is a clear jibe at the competition condition. What happens in real life is that policymakers - under the influence of bad ideas or bribe dollars - choose to protect and entrench the interest and market share of targeted (favoured) firms rather than have them compete. They also protected them from foreign competition via tariffs and import bans. What you get is a narrow set of beneficiaries with no incentive to improve their competence and production capacity. Import protection also mean they can collude on prices at the expense of the average consumer. The potential and reality of “state capture” has been the most important source of failure for IP in Africa. What is proposed is a systematic governance of markets. Rather what happens is succession of “control regimes” that ends up in higher prices, debt crises, and a lower average standard of living.

If we remove the influence of interest groups, can you count on policymakers to get IP right? You may be tempted to think that free of corruption, with ruthless execution, and intense monitoring, a honest government in Nigeria can get IP to work for industrialization. Such a hope is not wrong per se, it is simply an impossibly tall one. I have argued that governments are more susceptible to the so-called information problem because there are almost no direct costs for being wrong. There is no evidence that policymakers analyze the country’s competitive advantages when considering or investing in IP. There is always a strong preference for “sunset” industries that are considered once great or dying - a good example is CBN’s rekindled love for the good old great textile industry.

Economists usually study IP using cross-country regressions. They input data about a set of economic outcomes, (e.g jobs, income per capita, industry outputs) about a particular period of IP, for a selected set of countries, into regression models and then see what pops out. This method often miss a lot of the local contexts in respective countries being studied, but it is preferred because the findings can then be generalizable. There have been criticism of this methodology. Primarily because it tells you very little about the local conditions under which policies succeed or fail. And they cannot reliably inform you about causation - the precise reasons why some policies succeed and others fail.

There has been a resurgence of other approaches - mainly micro-economic econometric models. Think of it like a mathematically fancy version of case studies in business and management literature. Recently, economist Nathan Lane revisited the South Korea IP era using this micro approach. It is important to identify some of the “endogenous” factors in East Asia preceding the industrialization era. There was a lot of western involvement in the region due to the second World War and the Cold War after that. Situating the Korean experience against this background showed the prompt for its most aggressive period of IP. The Korean war already partitioned the country along North and South - with the U.S.S.R and the United States backing the respective sides. But American withdrawal of military personnel from the South was a shocking wake-up call. South Korea felt vulnerable and weak - especially with continued support of the Soviets and Chinese for North Korea. This led President Park Chung-Hee to rally the nation behind a rapid investment in the so-called Heavy Chemicals and Industries (HCI) to modernize the country’s industries and defense capabilities. The political economy of external threat is one that incentivize a disciplined approach to IP and discourages state capture.

The outcome of the Korean natural experiment was that there was a significant output growth in the targeted sectors. The output surge also lowered prices in the targeted sectors and downstream sectors (industries that purchase from targeted sectors). Upstream sectors (industries that supply targeted sectors) were negatively affected because of import competition - i.e government allowed targeted sector to massively import inputs. South Korea’s HCI policy was temporary and some of it were rolled back after General Park assassination, but it had a lasting effect on manufacturing. Many of the successful beneficiaries remain star players in the economy for decades after.

The lesson for Nigeria and other African economies should be to embrace nuance in their approach to IP. There are plenty of landmines to be navigated when considering targeted policies as opposed to broader economic reforms. Industrial Policies are very expensive. East Asia had decades of debt overhang to prove this. Secondly, the potential for state capture (see here for a Kenyan example) is ever present. State capture is the latent possibility that any targeted policy will be skewed to the benefit of firms and individuals closer to the government. Thirdly, the incentives are out of whack. Many African states like Nigeria are natural resource exporters and faces no external threat. This makes it hard to motivate policies to deliver good outcomes. Government survival depends on rents from resource extraction which can always be used to purchase the support of local rival groups. Finally, the bureaucratic capacity to administer targeted policies is either missing or has declined overtime. It is advisable for such states to put effort into developing capacity and broader market reforms than poorly executed and costly targeted policies. Industrial Policy remains a tricky, delicate business. See you guys next week.