Whenever an industry runs into trouble — and especially when it starts hemorrhaging jobs — demands for support and subsidies are heard. But does having an industrial policy really make sense? According to Howard Pack, a professor of business and public policy at Wharton, an interventionist government policy generally plays a limited role in bringing about an improvement. In fact, in some cases, government interventions can lead to harmful results, he adds.

An edited transcript of the conversation follows.

Knowledge at Wharton: Howard, thank you so much for joining us today.

Howard Pack: Thank you for inviting me.

Knowledge at Wharton: I wonder if we can start with an article from the Washington Post that you sent me some time ago. The article said that Air India obtained cheap loans funded by U.S. taxpayers to buy aircraft, which it then used to launch non-stop flights between New York City and Mumbai. This created jobs for the Boeing workers, of course, who made the planes. But Delta Airlines, which had offered non-stop flights to India, could not stand up to that competition and had to discontinue those flights, which presumably harmed other U.S. workers. Why does this happen? And what are the implications for industrial policy?

Pack: Let’s start in the beginning. “Industrial policy” is a set of policies which either provides subsidies … to firms which are currently producing, or protects them from foreign competition in one way or another. Usually the aim is to change the structure of production. In this case, it was to encourage the sale of Boeing planes. But as you point out, it has what economists usually call “general equilibrium effects” or “knock-on effects” — effects that hold a large number of implications for workers in other industries, as in this case, Delta. So the major effort here was to match stimulus that comes from other countries.

The U.S. is constantly trying to help Boeing relative to Airbus. Airbus has been the beneficiary of huge … subsidies from some of the European countries that participate in its production. The ultimate economic effects are very dubious, because, as you pointed out, the number of Delta employees decreased. On the other hand, the number of employees at Boeing went up. You’re never sure about the total impact. But employment should not be the only consideration. One consideration is clearly, what is the effect on the gross national product of the U.S.? This kind of subsidy usually has not had a beneficial effect in the long run on the U.S.

The political pressures are quite clear. Henry Jackson, who used to be senator from the state of Washington, was called the “Senator from Boeing.” He was a very effective senior senator, very good at protecting Boeing. Others are very good at protecting, say, the beet sugar industry. But most of the time it’s to the detriment of consumers and taxpayers. So what we basically see is a set of political trade-offs without much in the way of careful attention to the economic effects.

Knowledge at Wharton: We will come back in a bit to the economic effects. What are some of the traditional arguments in favor of having an industrial policy?

Pack: There are some good arguments for industrial policy, which would basically be something like the following: Say a company undertakes research and development. The R&D leads to unanticipated benefits for other sectors, which can then expand. There is then an argument for having R&D subsidies or perhaps tax benefits for R&D — assuming that you can encourage more, and that you’re not just rewarding existing R&D. So that would be one thing.

A second meritorious argument [can be made in this case]: Firms provide training for workers for which they’re not remunerated, and some of these workers go on to other firms. If the skills which the workers have acquired are of use to the other firms, [those] firms … are not going to pay the costs of the initial training. That would be an argument for training subsidies. In the lexicon of economists these are real external economies, and there’s an argument to be made to get involved in these.

But the real problem is that designing them very often becomes not a matter of economic analysis, but blunderbuss approaches. So, rather than giving an R&D subsidy or a training subsidy, there’s a demand for tariff protection for the firm or the industry, or generalized subsidies that don’t really target the specific goal you’re after…. Most often [industrial policy] is used to justify what are perceived to be improvements in employment opportunities, but very often these have relatively little merit and are very expensive for consumers.

Knowledge at Wharton: What does the empirical evidence show? Does it show that interventionist industrial policy makes sense?

Pack: I feel the empirical evidence is pretty clearly against it. [Interventionist industrial policy assumes] that the government — government officials — know the pattern of productivity growth in various sectors, which they can’t possibly know because the participants in the sector don’t usually know, and the government can’t predict these things. We know that that’s pretty much impossible. The usual examples of good industrial policy come largely from Japan in the 1960s, 1970s and 1980s, when it was asserted that the rapid growth of the Japanese economy was attributable to government intervention helping some sectors and not others. But the empirical evidence pretty overwhelmingly shows that the sectors that were targeted positively by the Japanese government were often sunset sectors, not sunrise sectors — sectors that were declining, which political forces tried to protect….

At the same time, in the 1980s, there was a fairly widespread demand [in the U.S.] for efforts to counter a Japanese resurgence in the semiconductor industry. It was beaten back by a variety of forces, and since then … the Japanese firms have basically had to abandon the field to either Korean firms or American firms, in this case Intel and AMD. So the evidence is pretty weak — and there has been a large amount of research on this — on whether industrial policy can be a positive force for most economies.

Knowledge at Wharton: As we know, the Japanese economy has been struggling for a while…. Why have they not been able to use industrial policy to revive the economy now?

Pack: The evidence is not very robust that they were able to use it effectively [before]. For example, it’s pretty well known to people who work on this stuff that the Japanese government discouraged its automakers in the 1950s and 1960s, saying, “You’ll never be able to compete with GM.” The sectors it did encourage [such as consumer electronics] probably would have done quite well without the encouragement of the government….

The whole point of industrial policy is that the government has to anticipate which sectors will emerge internationally. In its early stages, Japan had some easy targets. It was coming out of World War II — most of its productive capacity had been destroyed. But it has to be remembered that as early as 1905 Japan was able to mount its own battleships and defeat Russia in the Russo-Japanese War. So that basically in the immediate post-World War II period, Japan was targeting specifically American industries, most importantly iron and steel … in which it was able to borrow the technology, undertaking science/technology licensing agreements. That was easy because it was a stable industry with stable production and fairly high growth in the world markets. But to engage in industrial policy now would mean that Japanese officials would have to somehow figure out the innovation likely going on in the world, and compete with, say, Facebook…. That’s very unlikely and it would not generate that many jobs. And it is arguable that industrial policy actually has been partly, though not totally, responsible for Japan’s stagnation over the past two decades because Japanese banks do not have to evaluate loans. They were told by the government which sectors to lend to, and what happened was that once the country ran into a crisis the banks were more or less paralyzed in trying to figure out which firms to lend to.

Knowledge at Wharton: Like Japan, Europe also is struggling mightily right now. … Given all the challenges the Eurozone is facing, is there a case to be made that some form of industrial policy might help in securing the future of the Eurozone?

Pack: I think the eurozone’s problems really stem from a much different set of questions. Namely, the Eurozone assumed that by having one currency starting about 10 years ago, most [member] countries would fall into place. But one knew at the time that there were vastly different productivity levels between, on the one hand, the northern countries, like Germany and Austria and Holland, and the southern tier countries now referred to as PIG — Portugal, Italy, Greece and Spain, and one might throw in Ireland. I’m not so sure. If you take — to keep the example simple — the current problems in Greece, one of the problems is that the introduction of the euro allowed Greece to tap low-interest international loans. The Greek economy spent more than it was producing. … It imported more than it exported and was able to finance that by low-cost borrowing. If Greece were not part of the EU — say it were Senegal or Pakistan — it would go to the International Monetary Fund, and the IMF would tell it, “You have to reduce spending in the country to free up resources for exports. And simultaneously you have to devalue your currency so as to make your exporting more competitive” — or say, in the case of Greece, to make tourism more attractive to people from Northern Europe because it will be cheaper. But Greece is no longer on the drachma; it’s on the euro and therefore can’t do it. I think the problems in Europe are basically macro-economic in nature and not structural. [The structural problems that exist] tend to be rigidities in the labor market. So it’s not industrial policy one needs but a relaxation of some of the tight controls on labor mobility within the countries.

Knowledge at Wharton: Would you say that emerging markets like China, India and Brazil have been using industrial policy to some degree to drive their growth? What lessons can we learn from those markets?

Pack: Okay, so the simplest [example] is India. India has had remarkable growth since — one can pretty much date it — since 1991. India was in a very bad situation and asked the International Monetary Fund for guarantees and for loans. India accepted the IMF’s advice, which was partially formulated by Indian economists, that it should devalue its currency, take off quantitative restrictions on imports, reduce its tariffs and reduce industrial licensing. What India did then could be called reverse industrial policy — dismantling the industrial policy which had prevailed from 1947 until 1991. That helped India a lot. The runaway success stories in India in the IT sector — Infosys and WIPRO — were basically based on the fact that the IITs, the Indian Institutes of Technology, had produced large numbers of graduates who were very good and relatively inexpensive to employ. That had little to do with any government pro-IT policy. Indeed much of [the IT sector] was successful precisely because it could bypass the government. A satellite, which was really important for uploading and downloading programs, was put up by Texas Instruments, not the Indian government, although afterwards the Indian government ratified some of these things. In the case of India, [credit goes to] the dismantling of industrial policy rather than industrial policy per se.

Brazil, I think, is a different story. A lot of Brazilian growth has been based on agricultural production, commodity production. In some sense, it is critically dependent on exports to China — for example, soy beans. It does have an industrial sector but I defy you to go to Wal-Mart and find industrial products from Brazil. China is a different case. The Chinese government has obviously tried very hard to build up specific sectors. It has had [free-market-oriented] special economic zones, such as Shenzhen, and there, there has been some success. …

[Some] argue, correctly I think, that industrial policy has been partly responsible for China’s success, though its national investment rate of over 40% does contribute to it. But China is a very different story because it is so large that it is a very attractive target for foreign direct investment, and … that has helped propel its growth.

Knowledge at Wharton: A very interesting phenomenon is the growth of global supply chains and production networks. How do they influence industrial policy?

Pack: That’s a very important question actually, because many countries, especially in sub-Saharan Africa, but also in the Middle East and North Africa, have very small manufacturing bases. Most sub-Saharan Africa countries have a ratio of manufacturing employment to total employment of less than 10%. The same thing is true in the Middle East and North Africa. Having worked on both recently, my understanding of it is that these countries really have a need for foreign direct investment to provide jobs in the short term. The Arab Spring might have been precluded had there been more rapid FDI inflows into these countries. So becoming part of global [supply] chains is very important. It’s one of the few quick ways into manufacturing because the firms that locate there will pay for the investment, bring in the equipment, bring in management, and so on and provide a lot of jobs in the short term…. Firms look all around the world. Currently wages in China are going up pretty rapidly, and it is possible some of the jobs that China will vacate in the low end of the spectrum will migrate to the Middle East and North Africa and sub-Saharan Africa. But the evidence is [that the work is] going to Cambodia and Laos and other places in Asia because they have pretty disciplined and diligent labor forces and they’re close enough to China so that they can interact a lot with Chinese suppliers or intermediate inputs…. So far, the results do not suggest that many firms are interested in [locating plants in Africa or the Middle East].

Knowledge at Wharton: One final question: You’ve made a very compelling case that the results from industrial policy are quite mixed, and there are strong grounds for skepticism. Again, instead of being interventionist, in the absence of industrial policy, what can be done to deal with market failures?

Pack: For the kinds of things I was talking about earlier — namely, training takes place, firms pay for it, and then other firms benefit from it — you do need tax and subsidy policies. But you have to identify where the failure is and address the specific failure…. That is really different from a generalized, blunderbuss industrial policy where you say, “Let’s protect an entire industry,” rather than trying to target the particular sources of market failure.

Knowledge at Wharton: Howard, thank you so much for speaking with us today.

Pack: You’re welcome.