Sunday, May 27, 2012

Chang’s (2003) argument about role of state in developing economies.

• 1. Introduction

There is much controversy over the government’s role in developing the country’s economy, views ranging from developmental state and laissez-faire. Since the failure of several nations (particularly in Africa) to develop despite extensive government involvement, more market-friendly approaches have been advocated and preached for the last couple of decades, yet there is little sign that this has the ability to move countries up the ladder either, and so the debate continues.

It will be argued in this essay that industrial policies have worked and that there is a case for industrial policies and an active government role in the development of a country. Yet it will be recognised that there are problems that will have to be overcome. The issues at hand are not simple and straightforward.

Firstly, the reader will be introduced to what industrial policy is and what the different policy tools that have been used are. The second section will look at the empirical evidence for and against industrial policies working as a catalyst for economic development, and inspect the problems of market failure, in short the section will examine the case for industrial policy. Finally, the question about government failure will be raised. Is the government able to solve market failures despite its problems?

2. What is industrial policy?

Before any debate on industrial policy can start, defining the concept is of importance. Chang (1994:59-60) discusses several definitions, finding them too broad and general he settles down for the much narrower definition that industrial policy. is “a policy aimed at particular industries (and firms as their components) to achieve the outcomes that are perceived by the state to be efficient for the economy as a whole” (Chang, 1994:60).

“Perceived” needs to be emphasized as this gives room for the failure of governments if there is a gap between the perceived and the actual efficient outcomes, and “the economy as a whole” means that there is room for industrial policies to reduce efficiency in some area of the economy at some time, if it is perceived that there is/will be a net gain in the economy as a whole.

The definition given by Chang still allows for a wide variety of industrial policies. Just to give the reader a taste of the wide range of policy tools available, some few of the means by which states have used industrial policy are listed below:

• Tariffs (import substitution) and taxes
• Local content requirements (ownership and inputs)
• Discriminating interest-rates
• Direct subsidies
• Investment guarantees
• Sector-specific skills development and R&D

3. Why Industrial Policy?

3.1 History and Empirics: Has it ever worked?

Considering the heated debate surrounding the viability of industrial policies as a tool to achieve industrialization, a natural starting point would be to examine history in order to uncover the empirics of success and failure of industrial policies. Have such policies ever been able to assist development?
Lal (1983:71) claims that there have been no countries that have achieved higher growth as a result of industrial policies. Countries having employed industrial policies and developed have rather done so despite of industrial policies implemented. He goes further and states that countries that maintained free trade regimes have consistently outperformed those who intervene in markets through the use of industrial policies. (Lal:1983:71). However, Lal fails to give one single example and thus his claims are not substantiated.

Along the same lines, Krueger (1990:12) argues that there are no evidence that living standards were falling in developing countries prior to 1950 and that the deterioration happened post 1950’s when government intervention was high. From this she concludes that intervention in the market is a hindrance to development.

Note however, that from the statement “there is no evidence that living standards were falling” it does not logically follow that living standards were actually not falling. The lack of evidence might just be due to the lack of data, reliable data is notoriously hard to get in developing countries even in today’s world. In addition to this, Krueger does not present any evidence that living standards were rising either in the period prior to 1950.

Furthermore, to claim that developing countries were in the utopia of laissez-faire prior to 1950 is a crude statement considering the fact that several (in fact almost all African) countries were under the rule of colonisation during that time, subject to severe trade-restrictions and often exploited by monopolies from the “mother countries”.

Despite the weakness in the above claims made by Krueger and Lal, there is some truth in saying that industrial policies have failed in developing the economies of many countries, in particular African countries.

Import-substitution and trade taxes amongst other interventions were heavily used in African countries with disastrous results. It becomes clear that industrial policy, like text-book free-market economics, is not all that simple in reality. There are however cases where industrial policies have been implemented and the desired results were achieved.

Chang (2003), in direct contradiction to Lal claim that all the developed countries preaching free-trade today were built upon industrial policies. In his book “Kicking away the ladder”, Chang extensively discuss several case-studies and presenting data on previous tariff levels of the then industrialising countries, and it is worth noting the highly involved governments of the United Kingdom and the USA.

More recent famous cases of countries successfully implementing industrial policies are Japan and Korea with their governments’ close collaboration with the industrial entrepreneurs, and Norway’s local content requirement with respect to the oil-industry, spurring national research and technological innovation in this field, moving the country up to higher technological production.

Compared to Lal’s sweeping statement, Chang’s detailed investigation to the former policies of developed countries should be given weight. It should at least be accepted that industrial policies have worked and can work in some instances. The question at hand is thus not whether the use of industrial policies are good or bad, but rather what industrial policies are needed for success.

3.2. Theoretical Justifications and market failures:

This section will mainly draw on the ideas from Chang (1994) and Rodrik (2004) about the failures inherent in the market and thus the necessity of government intervention.

3.2.1. Coordination problems of the free-market economy

The kernel of this type of failure is that ex-ante coordination problems create efficiency losses to the economy.

In the perfect decentralized model it is assumed that firms can enter and exit markets without incurring a non-recoverable loss (waste) and that there is no interdependence amongst firms. If there are profits to be made, firms will enter the industry, perhaps more than what is viable, but this will only mean that the least productive firms will go bankrupt and leave the industry, leaving the market to clear at the most efficient outcome. The market corrects itself ex-post, so there is no need to correct for coordination failure (excess firms entering the industry) ex-ante by the government.

In real life however, investments in a bankrupt firm cannot be “instantaneously and costlessly shifted to other activities” (Chang: 1994:65). Due to this fact, the markets ex-post correcting mechanism for the coordination failure is sub-optimal, since this will incur a waste in the form of “asset specificity”, i.e. investments made into assets that cannot be shifted to other industries. Only ex-ante correction of the coordination failure will ensure that there is no waste incurred by bankrupt firms. It is for this reason government intervention is needed in order to ensure that there an optimal number of firms entering the industry.

Investment coordination is needed to avoid both over-investment and under-investment. Due to the fact that firms know they may face irrecoverable losses in terms of specific assets, they may be too risk weary to enter an industry, and this could lead to under-investment in the industry. If there are too many firms entering the industry, competition will force some firms to exit, thus resulting in their specific investments being wasted. The government can in these circumstances facilitate and settle private bargaining among potential entrants. Another method to coordinate entry is to implement “conditional entry”. By this it is meant that new potential entrants will only be allowed to enter the industry if there are increases in demand, so called production licenses will have to be given (Chang:2002:66-67).

If demand was to decrease, thus shrinking the number of optimal firms in the industry, the government needs to step in to ensure that specific asset investments are not going to waste. This has been done in Korea and Japan by the government creating recession cartels (during temporary falls in demand) and negotiated exit (for more permanent recessions).
In a recession cartel, the firms agree to reduce their supply for some time, thus allowing all the firms to stay in business, and hence no exit-waste. These types of cartels are justifiable since the more exit there is of firms the greater power the firms left will have, and it could even lead to a monopoly situation (Chang: 1994: 67).

In some instances, there is a need for firms to exit even if it causes some waste, such as if the fall in demand is permanent. However, if some firms exit (incurring a loss), the remaining firms will profit. This will lead to no firm wanting to be the first to leave the industry. This dragging out of the inevitable scenario results in everyone being worse off. The excess supply may cause the price to plummet, and all firms may be able to go on for some time at a loss, but eventually some firms will go bankrupt being forced to exit and thus still incur a loss of investment. If exit must happen, it should happen as soon as possible. The government can arrange compromises between the firms, in order to generate voluntary exit from the industry, for example by ensuring that the remaining firms pay compensation to them (Chang:1994:69).

3.2.2 Vertical Investment Coordination
When it comes to investing in new a new industry, although it may be potentially profitable, entrepreneurs may be unwilling to enter the industry if the profits depend on investments being made by other actors. It can be illustrated as an assurance game, where all actors along the vertical chain wish to invest if the others do so, and not to do so if others do not (as the profits depend on all investments being made). There are two equilibria in this game, where the Pareto optimal is [invest, invest], yet the economy is often stuck in the lower equilibrium where no-one invests, since there is a clear first-mover disadvantage.

Here the government has a vital role to play. It can create a focal point by announcing plans and encourage the different actors to invest, perhaps by offering preferential interest-rates on loans for investment in the desired areas (Chaudhuri: 1990: 32). A key characteristic of the game described above is that the risks and costs are borne by the individual but the benefits generated are enjoyed by others too. Thus socialisation of the risks, such as investment guarantees (where the government guarantees to cover losses of investment of a firm who takes the initial step, if other firms don’t follow suit and profits are not realised) is a good way of inducing investments (Rodrik: 2004: 14). If the government does not intervene in these circumstances, economic change may never happen (Chang: 1994: 76-79).

The example of the Taiwanese orchid industry given by Rodrik (2004) would serve well here. The orchid industry has proved to be profitable, however it is doubtful that they would have ventured into this area if the government had not ensured that all complimentary investments had been made, such as the availability of irrigation, logistics and transport networks, electrical grids nearby, quarantine, marketing abroad and so on (Rodrik:2004:12).

Another problem with investments into new industries, like the one described above, is that it is unknown territory. With no previous prices and costs known to entrepreneurs, it is not possible to know ex-ante whether a particular industry is going to be in a country’s comparative advantage. Thus, without being able to refer to costs and prices, this uncertainty will deter investment into un-ventured industries. The government must be involved in a process Rodrik refers to as “self-discovery” (Rodrik:2004:12). If one industry proves to be unprofitable, the resources devoted to the “experiment” have not been a waste, since self-discovery is not merely about finding profitable industries. Rather, self-discovery is a learning process aimed at increasing the level of information for entrepreneurs. Hence the government must provide support for entrepreneurs to enter industries with unknown parameters so that they can fully exploit all potentially successful industries.

4. Perceived problems with industrial policies:
There are several problems associated with industrial policies, but due to the limitations of this paper, only the main and most serious criticism will be dealt with in this section, as well as responses to these objections.

4.1. Information failures
In a sense the problem of information failure can be divided into two categories:
4.1.1. Insufficient information.

This objection to industrial policy usually argues that the government will not be able to overcome market failure as they posses no more information than economic agents. If the government does not posses perfect information then they will not be able to plan correctly, hence the industrial policies implemented will be inaccurate and not induce efficiency gains.

However, as Chang (1994: 80) writes, having insufficient information is not really an obstacle to planning, uncertainty is the core reason for why we try to plan in the first place. If everyone had perfect information, planning would be redundant.

Furthermore, considering that business do not possess perfect information either, it is hard to see why this argument (of government possessing insufficient information) gives weight to the statement that businesses should be in charge of planning. At the most, it could be argued that neither government nor business are better than one each other in planning for the future, but not that business are better equipped than the government to do so.

4.1.2. Asymmetric information.

The argument against industrial policy based on the belief that there is asymmetric information between firms and government and that this lack of “local information” possessed by firms cause the government to be less able to make appropriate decisions regarding the efficiency of the economy.
Again, Chang replies to this that asymmetric information is a problem that everyone has to deal with, between the government and firms, amongst firms and even within firms. Yet economic agents can function and overcome these obstacles, there is no reason why the government should not be able to do so either. Furthermore, this problem persists in all areas of activity between any agents, so if one wish to believe that this argument is strong enough to dismiss industrial policy then one must also be prepared to dismiss all sorts of policy, as other areas will also be suffering from the problem of asymmetric information (Chang:2002:82).

4.2. Government failures:

4.2.1 Rent-seeking, corruption and weak political will.
A very common line of criticism towards industrial policy and government intervention is the propensity for these to generate rent-seeking and making the government and its bureaucrats prone to corruption. A country with weak political will or with underpaid bureaucrats and politicians may not be able to withstand the pressure from industries’ interest groups to prolong protection or other (intended temporary) support offered by the industrial policies (Pack: 2000: 64). This would result in the industries remaining inefficient and never build their competitive edge up as preached by developmental states. The result is a perpetuated drain in resources, a waste and lack of an efficient industry.

All though the point above is valid, it is important to realise that the problem here is not with the industrial policies, but with other political factors. The conclusion should not be to avoid industrial policies, but rather to emphasize and build strong institutions and regulatory system of punishment and rewards. As Chang (1994:84) puts it, the state needs to set a strict set of performance criteria that needs to be met, and the state needs to have a punishment mechanism available to deter firms resisting the termination of certain policies intended to be temporary. If the government can make such threats credible, firms will not slack off during the period of protection, but instead bolster their competitive edge up as much as possible (which is the ultimate goal of industrial policies).

Furthermore, although protecting the industry from excessive competition from abroad, it does not mean that all competition is banned from the industry. The state could ensure that there are enough domestic firms to generate moderate healthy competition, perhaps allowing the firms to compete for future protection, where the government will provide support for a group of best performing individual firms rather than the industry as a whole.

The situation of rents in the case of industrial policy is similar to that of rents accrued to a patent holder. In the patent holder case there are often strong set of rules dictating how long the patent holder is protected that are rigid. Institutions preventing excessively prolonged protection as a result of successful lobbying by interest groups could be built using the patent laws as a model. Patents are even promoted by neo-classical economist as innovation is important, if it is possible to create a balance between incentives and restricting monopoly power in the patent case, then there is no reason to believe this can’t be done in the case of industrial policies either.

Lastly, as Rodrik (2004:37) points out, rent-seeking behaviour can occur with any type of policy, even those regarding macro-economic fundamentals such as privatisation, and is thus not restricted to industrial policy, yet there is much reluctance to accept this. One can thus not argue that industrial policies must go out the window while maintaining that the government has a role in creating other necessary macro-economic policies.

4.2.2. Comparative advantage of the government

Krueger (1990:17) argues that the government has a comparative advantage in providing and devoting resources towards macro-economic fundamental policies (such as building schools, improving infrastructure and communication amongst other things), and should not devote its time allocating investment in terms of industrial policy. If you have a comparative advantage in something, then you spend your resources more efficiently doing that than anything else, thus the government is generating inefficiency when committing itself to industrial policy as it distracts resources from where they would be most effective: macro-economic fundamentals in accordance with the Washington Consensus.

Krueger’s argument is logical, yet the premises on which it is built are weak. What is it that makes the government have a comparative advantage in dealing with macro-economic fundamental policies more than it can deal with industrial polices? No explanation or evidence is given for this claim, and there is no reason to believe that the government is more competent in investing in infrastructure and education than it is in investing in its other industries, so the argument above cannot be considered a valid criticism towards industrial policy.

5. Conclusion
At the end of this paper, we should summarize the questions raised and look at the answers we have found. Do industrial policies work? History tells the tale of many successful uses of industrial policy, yet also of some failed attempts. It is fair to say that industrial policy can work.

The theoretical insights made by Chang (1994) show that the market posses some inherent flaws that need to be resolved in some way or another. In the face of the problems inhibiting today’s developing nations, is it plausible to believe that they can overcome the government failures so to be able to use their state as an effective policy maker with the interests of the economy in mind? Some countries managed to do so, others have not. This highlights the significance of the context in which the policies are implemented, as policies need to be shaped accordingly.

It is at this stage important to ask the question -What are our options? Should we abandon industrial policy (and all other policy making for that sake) if we run into challenges, and turn to the just “do nothing” option and let a flawed market take care of things? Krueger (1990) and Lal (1983) both claim that government failures are more sever than market failures, yet they both omit to explain or prove why this is always so. It is not implausible to believe that the magnitude of the two failures will differ from case to case? Too often, academics decide on an ideology or position they want to hold a priori, and then look for arguments supporting, instead of taking a long good look at the evidence at hand and then decide on the opinion to hold. The neo-liberal position is filled with an irrational state-phobia.

The new theoretical work on market failure tells us there is a case for industrial policy and government intervention. History and empirics tells us it has been done successfully. Industrial policy can thus be viewed as a necessary but not sufficient condition for achieving economic development and efficiency.

Instead of giving up all together on industrial policy when it runs into problems, lets recognize the need for more objective research on the topic, focused on the conditions that make industrial policy work and how these can be achieved, rather than engage in an endless debate on whether it is a “growth elixir or poison” (Pack:2000).


1. Chang, 1994 “The political economy of industrial policies”

2. Chang, 2003. Kicking away the ladder, London, Anthem.

3. Chaudhuri, M. 1990. “Market failure and government failure” in Journal of Economic Perspectives vol.4 no.3

4. Kreuger, A. 1990. “Government failure in development” in Journal of Economic Perspectives vol. 4 no. 3

5. Lal, 1983. “Industrialisation and Planning”

6. Pack, H. 2000 “Industrial Policy: growth elixir or poison?” in The World Bank Observer vol. 15 no.1

7. Rodrik, D. 2004. Industrial Policy for the 21st Century