Democracy and the Future of Ethiopia’s Developmental State Conference
Professor Francis Fukuyama, Stanford University
Hyatt Regency Hotel
Addis Ababa, Ethiopia
June 11, 2019
Thank you for the warm and generous introduction. I am really delighted to be here, in Addis, today. This is my first visit to Ethiopia. This is a country that I have followed over the decades, with a good deal of interest, and it is very satisfying to actually be here and talk to people about the remarkable changes that are going on in the country. My talk will be about developmental states, because this is a subject that I have been interested in over the years. I don’t think I can give Ethiopia much specific advice, but I do think that an understanding of the way developmental states have, and have not, worked in other parts of the world is important.
I actually, about 15 years ago, was given a copy of Meles Zenawi’s Master’s Thesis, that he wrote as a correspondence student for a Dutch University, which was an extended critique, basically, of the reigning neoliberal development paradigm. I remember reading it back then and being absolutely astonished at how somebody that was actively a commander in a civil war could take the time to study and to write as sophisticated a thesis as this one. I think, it very clearly laid out the theoretical basis for the Ethiopian developmental state that would follow.
Now, the history of thinking on developmental states has had its ups and downs. It was very much opposed by most free market, American economists back in the 1990s. After the Asian Financial Crisis, and up until the subprime crisis in 2008, I think there was a feeling among many economists that, even if developmental states had worked in Japan, South Korea and China, that they were becoming outmoded. But I think that the crisis, and the rethinking of development that happened after 2008, has led to a resurgence of interest in the developmental state, and a degree of respectability. So, if you look at some of the democratic candidates, like Elizabeth Warren, that are trying to run for president next year, the developmental state or a more active state is definitely on the agenda.
So, I want to just go over a little bit of the theory behind the developmental state, and then what some of the pitfalls and advantages are.
The critique of the neoliberal model, I think, is pretty well established at this point. The Washington Consensus ‒ however you want to label that group of free market policies ‒ made some fairly heroic assumptions about economic conditions that were faced by developing countries. Most important, was it assumed that institutions existed, property rights existed, the rule of law could adjudicate disputes, and it assumed that there was going to be perfect information on the part of economic actors about market opportunities. It also assumed that financial systems existed, transaction costs were very low, there would be no external shocks, and the international environment would be permissive of development if you just simply get the state out of the way. And of course, this was wildly unrealistic for most developing countries.
And in fact, the theory behind developmental states made a lot of sense in economic terms. For example, infant industry protection was one of the key pillars of many developmental states. It makes sense. The United States itself practiced a high-level protection of its manufacturing industry sector throughout 19th century as it was becoming internationally competitive. The idea is that it takes time to scale up, to walk down a learning curve, and to be ultimately internationally competitive.
Similarly, I think that, and this is a point that Dani Rodrik has made, there is a huge coordinating function that states can play very usefully in the development process. If you want to build a factory, you are not going to invest in it unless you know that there is electricity, roads and rail roads to take your goods to a port. All of these complementary institutions and investments have to be made before that private sector actor is going to make the investment in the factory, and the coordination that is required to ensure that all of these different pieces exist and come together, is something that oftentimes a state can do, primarily through its investments in things like infrastructure. In fact, this was the kind of policy that many East Asian Developmental States undertook. It was a series of coordination problems, one after another, that occurred in the process of development where the state can actually play a helpful role. And finally, I would say, in favor of the developmental state, that if you are developing late, meaning you are not the first country to industrialize, it is easier. You do not have to invent technology, you can buy the technology. You can see what the next stage in the industrialization ladder is going to be, and you can target moving up that ladder. This is exactly what a state like South Korea did.
The critique of these developmental state theories was made by a lot of economists. Paradoxically, the critique is more political than economic. The critique really had to do with rent seeking. And the argument was that if the government put itself in a position where it was directing investment through subsidized credits, export licensing, or other forms of state intervention. It was a very powerful tool that would be used by politicians not to pursue developmental ends, but to line their own pockets. That infant industry protection, for example, while it worked well in certain countries like South Korea or Japan, it did not work so well in Pakistan, Argentina or many other places, where the government wants the infant grown up to a viable adult, it could compete globally; but the government did not take away the subsidies. They did not allow the child become an adult, and therefore created non-competitive protected industries for essentially political reasons.
And the state, in any event, would not know how to pick winners, it did not have the foresight. Especially as the level of technology began to approach the global frontier of technology development, the state would be much less capable than the private sector of actually knowing what the next stage of technology would be. So, the Japanese industrial state ran into a big problem in the 1990s with its so called fifth generation computer program – it had no idea that personal computer was going to be the frontier of computer development. The bureaucrats did not have any better idea than private market participants.
This was the argument that has gone back and forth on the advantages and disadvantages of developmental states.
If you look at the academic literature on the developmental state, I do think that there was consensus on a certain set of conditions that really had to be met, and were met, by many of the East Asian fast developers for this kind of state-directed development to work.
By the way, there is a very nice little book published last year by Stephan Haggard, professor at the University of California, San Diego, called “Developmental States”. I strongly recommend it. He was one of the longtime observers of this phenomenon. And he has encapsulated this entire debate in a very compact book – it is only about 100 pages. But I would say that the conclusion of the scholarship says the following: that if a developmental state is actually going to succeed the way that the ones in East Asia did, the single most important condition is state quality. If you do not have a state that has a certain set of characteristics, it is probably better not to pursue this in the first place because the state is, basically, going to screw-up the process.
What are those qualities?
First of all, you have to have a capable bureaucracy. This was Chalmers Johnson’s argument in Ministry of International Trade and Industry (MITI) and the Japanese Miracle. MITI was an example of East Asian highly professionalized, highly technocratic bureaucracy. It was recruited based on meritocratic grounds. And people were promoted on the basis of achievement.
One of the important characteristics of this kind of bureaucracy was what Peter Evans called “embedded autonomy”. What that meant was, autonomy from short-range political forces. That is to say, a politician could not go to that bureaucracy and say, “hey, by the way, my cousin has a factory and it needs help, so I want you to subsidize a loan to give my cousin that kind of assistance”. The bureaucracy was sealed off from that kind of short-term political influence, but it was embedded. What embedded meant was not that they are a bunch of bureaucrats sitting at sealed offices completely separate from a society over which they are presiding. They were, in fact, in constant dialogue with the private sector - with companies, banks and investors that had on the ground information about, for example, those sorts of coordination problems mentioned above. They were open to dialogue, including with civil society groups, to find out what the needs of their society were. So, they were aware and responsive to the needs of their societies. That is also to say that they are accountable, but they were not subject to short-term politicization, which affected bureaucracies in Latin America, the Middle East, and in many parts of the world that tried to do this kind of industrial policy. And it is very interesting, I believe, frankly, that there is a cultural dimension to the ability to run an effective industrial policy, and I don’t think it is an accident that so many of the successful countries that have run industrial policies are all located in East Asia.
The final, really important, condition for an industrial policy to work, is a kind of moral attitude on the part of leaders. The people at the top of the hierarchy that are directing the bureaucracy have to believe in development. They have to believe in national development rather than lining their pockets or the pockets of their families.
For example, in South Korea, Park Chung-hee staged a coup in 1961, he replaced a very corrupt administration that had been in power for a previous decade. He was an authoritarian leader. He disciplined the large Korean Companies, Chaebol, like Hyundai and Samsung. He actually put dunce caps on their heads, and paraded them down the streets of Seoul and basically told them that they had to exert discipline over themselves, if the government is not to put them in jail. And what that discipline meant was reinvesting 70% of their earnings back in the Korean economy.
In China, the initial phases of their economic growth after 1978 (after the passage of the Household Responsibility System, which, in effect, created a form of private property in agricultural land) was built based on something called township and village associations. These associations were run by local party officials, and under the discipline of highly centralized and pretty authoritarian communist party, these officials were told: “You can go out and make money. You can invest your local assets in a casino, or in a factory or in a tourist agency or whatever you want,and you can actually keep some of that surplus for yourselves. But again, you have to plough back two thirds of the earnings, from whatever activity you are engaged in, back into your province or city or locality. And if you are not going to follow that instruction, you are going to go to jail.”
And this, in a sense, is why, I think, that these societies manage to run these policies effectively than in other countries. If you compare these experiences with let us say the cases in Kenya, Nigeria or Indonesia or other countries, you would have had leaders that took different positions, like saying, may be, we will invest 5% and the other 70%, we will go pocket it in a Swiss bank account.
So, there is something about the public facing moral qualities of the people leading these societies that was very critical.
I must say that the two African countries that had been singled out and compared to the East Asian developmental states – Ethiopia and Rwanda – do have something in common in this respect because the people that put their industrial policies into effect, I think, also had the sense of discipline and understand the need for the state to discipline the society in order to prevent the most egregious forms of corruption and rent seeking. And that was important.
But the question, I think, all of this raises is how long can a hierarchical party, that is not subject to democratic accountability, maintain that kind of a posture? How long can it, in effect, remain disciplined and not be susceptible to demands from party members, from the society and from private interest of various sorts for the kinds of rents that the state has at its disposal?
Those are the conditions, I think, and are fairly strict ones. And that is, by common agreement, what has made East Asian countries successful.
One further point, which may be added here, has to do with issues of social protection. Many people would argue that the East Asian developmental states were authoritarian and that authoritarian power is necessary to their success, in part, because they suppress demands from labor. This was true especially in South Korea and to some extent in Taiwan, where Labor Unions were banned after the dictatorship came into power. They succeeded in holding labor cost down until the moment, that in 1987, for example, Korea democratized. I think that more careful study of what these governments did made them not as quite harsh as that, because they actually were listening to social demands. And even under the authoritarian governments, they did put into place, for example, healthcare policies, invested very heavily in education and other kinds of social services, because they understood that this was also part of a broader emphasis on development.
Let me know turn into a comparison of the ways in which Ethiopia does and does not compare to these East Asian models. In many respects, Meles took South Korea to be a model for what he wanted Ethiopia to be and I count five major respects in which Ethiopia differs.
The first respect is that the Asian developmental states all started with a very vigorous and entrepreneurial private sector. In fact, they all started with a private sector that had companies that were nascently competitive to the point where the government felt it needed to really discipline those companies and show them the sense of who is boss. But it really was a partnership and not state-domination – it really was the state learning from the private sector, helping the private sector, and getting feedback from the private sector because the state wanted the private sector to be the primary driver of development. This is true, including in China, where you did have a lot of state-owned enterprises that remain the national champions of that economy, but they are much less productive than private Chinese companies that the Chinese government has encouraged very strongly.
Ethiopia, as far as I can tell, does not seem to have much of a private sector. Ever since I have gotten here, I have been asking people what percentage of GDP or what percentage of aggregate output is produced by private firms and nobody seems to have a clear idea of that. But everybody agrees that the number is pretty small.
So there has not been a partnership between strong government and strong private sector. It has basically been state-driven development with very weak private sector, and going forward that is going to be a real disadvantage.
Second difference, the Asian developmental states had modern bureaucracies in place at the beginning of their big developmental push. A little bit different in China. China has the oldest national tradition of meritocratic bureaucracy. It really is something that has gotten its start in Qin unification of the 3rd Century B.C. In China’s case, there is an interaction between the party and the bureaucracy. The party is a very hierarchical and disciplined party that recruits its cadres on the basis of merit. You have to take an exam in order to get into the Chinese Communist Party (CCP).
So all of these countries had these kinds of bureaucracies with embedded autonomy, high level of technocratic capacity based on good education (MITI and the Finance Ministry of Japan has been filled with graduates of the Tokyo University’s Faculty of Law, which is the most elite education institution in the country) and very little politicization of those bureaucracies. And this is not something that is true to Ethiopia. The civil service of Ethiopia is not of that quality.
The third big difference has to do with what I would regard to as a problematic relationship with China. I don’t participate in certain kind of China bashing because China builds a lot of infrastructure. The West has largely given up on building infrastructure. It has not been very helpful to developing countries as they need electricity, roads, rail roads and all sorts of infrastructure. So I don’t contest that.
But a lot of Chinese investment comes in a big package where the due diligence the Chinese perform in these investment is not terribly good, in terms of estimating accurately internal rate of return (IRR) and externalities. I think that this has led to a situation of over-indebtedness. Because the Chinese demand sovereign guarantees for their loans, virtually all of the risk of a failing project is on the country that is the recipient of that investment. Those risks are not born by Chinese companies and Chinese financial institutions.
The fourth issue has to do with the international environment. It is just more difficult today to be a developing country that it was when South Korea, Taiwan, Hong Kong were in their high growth phases. That period, we were in the middle of a Cold War in which the US, allowed considerable free riding of its own domestic market by these East Asian countries, and allowed them to run trade surpluses over extended period of time. This is something that has produced a big backlash in the US. That is why you have Donald Trump as president today. But this is a phenomenon that is not just in the US. In general, the willingness of developed countries to take exports from developing countries is less.
And the other big issue, I think, is what sometimes referred to as premature deindustrialization. It is very hard for developing countries to break into an existing form of manufacturing because the efficiency of East Asian producers is so great; we don’t need another automobile company in the world today. The world is awash in too many automobiles as it is. So the path that South Korea followed, I don’t think will be followed by any other country in the future. There has to be a different sequence of steps and what that is quite uncertain.
The final issue has to do with state building and nation building. This is where identity issue comes in. Japan, South Korea, China and to some extent Taiwan all had well established states and well-established national identities before they began their modernization process. In fact, my view is that the reason that they are spectacularly successful is that all of those issues were settled hundreds of years ago. Common language, common ethnicity, common historical traditions and common bureaucracy were all in place before they confronted Western imperialism in the 19th Century. Therefore, after the end of that imperialism, when they were left on their own, they already had state systems and national identities that could be the basis for economic takeoff. But this is not the situation of Ethiopia, where national identity is contested.
In fact, because the nature of the power distribution and settlement that came about in the 1990s, it has kind of embedded identity politics into the constitution itself. And once that happens, like the former Soviet Union, it is very hard to get out of it.
Even with the EPRDF, if you compare it with the CPC of China, it is a different kind of party. It is not centralized. It may be disciplined, and has been disciplined in earlier phases, but right now it is essentially an ethnic coalition and therefore cannot be used as an instrument like the CPC could be used for.
So just to conclude, I would say the major things seems to be lacking in terms of Ethiopia’s developmental states are:
· Ethiopia needs a modern state. It needs a bureaucracy that is meritocratic, technocratic, high capacity, and insulated from politicization.
· Ethiopia needs a unifying identity. Coming to a country like this, the last thing I want to do is give you advise on a sensitive topic like this. But one of the bottom lines of my last book is that if you have a highly diverse society, you need to decentralize important forms of political power and give people local autonomy. But that can to coexist with an overarching sense of common nationhood in terms of the stories and symbols that people hold such that they actually believe that they are part of a larger enterprise. It is extremely hard to achieve in a country as diverse as Ethiopia but I think that is necessary.
· Finally, Ethiopia needs democracy. You know, one of the bad consequences of developmental state is its concentration of power in an executive. That was necessary for the developmental aims that Ethiopia was seeking, but at a certain point it becomes counterproductive. It is basically trying to suppress social forces that are active on the ground and that are not happy with that kind of dictatorship, they are going to come out sooner or later. China may look stable and strong at the present moment, but believe me there is a lot of social forces that are simply being suppressed and at some point it is going to come out.