I first encountered Robert Atkinson when he was working at the Office of Technology Assessment on Capitol Hill in the early 1990s. OTA was an invaluable resource for anyone who wanted to understand American industry, and when Republicans won back the Congress in November 1994, they naturally shut it down. Rob subsequently went to the Progressive Policy Institute as its vice president, and then in 2006, he founded the non-partisan Information Technology and Innovation Foundation, which does a lot of what OTA used to do. During Barack Obama’s presidency, it was the place to go for analysis of the high-tech industry, and also a prime promoter of industrial policy. ITIF helped inspire the administration’s attempts to set up programs in advanced manufacturing. I wanted to ask Rob what he thought about Donald Trump’s pledge to revive American manufacturing, which has been widely criticized by liberal and conservative pundits. Is Trump making good on it or not?
Judis: Let’s leave aside, for the moment, the question of what Trump is actually doing. What do you think of his pledge to bring manufacturing back to the United States?
Atkinson: First of all, it’s a worthwhile objective to grow manufacturing in the United States and create more manufacturing jobs. But you can’t do that only by bringing manufacturing back. You’re going to have to have a combination of bringing some jobs back and expanding manufacturing exports more.
Judis: But it is feasible to do that?
Atkinson: It’s absolutely feasible. There is no inherent reason the U.S. has to be running around a $600 million trade deficit in goods. We could be neutral. We could even be running a trade surplus in manufactured goods. If we had the right policies, we could do that. So it is certainly feasible. The risk is if people start believing that we are going to be bringing back T-shirt production or plastic injection-molded toys they give you in your fast food restaurant bag. We are not bringing those back, and frankly they are not very good jobs. But there are some jobs that are better and that we could be bringing back. And indeed there is some reshoring going on now.
Judis: What kind of industries could come back?
Atkinson: Some kind of industries that can come back are where the goods weigh more, and have higher transportation costs. Also industries that produce complex goods that require higher skills and technology to make them, areas we have some advantages in. For example, this is why General Electric brought back appliance production from China and Mexico to Louisville, Kentucky.
Those are the kind of things that can be brought back, but it is very unlikely that products like iPhones will be assembled in the United States. iPhone assembly is very routine and low skill, which is why it is done in China with its very low wages. The only way you could bring that work back would be to highly automate it, but that is very hard to do with today’s technology. So the type of work that is likely to be brought back would be more skill intensive or more easily automatable.
It does take a fair amount of skill to assemble a car or to make drugs in a pharmaceutical factory. It is harder for someone with a sixth-grade education to do this work. So it is the more technical work we could bring back and where there could be higher levels of automation to make the higher U.S. wages competitive with low-wage foreign labor. Wages in Mexico and China are only around 15-20 percent of U.S. wage levels. That doesn’t mean everything goes there, because one of the reasons their wages are lower is because their productivity is lower. So in areas where U.S. productivity can be high, the work can be brought back.
Judis: One of the objections to Trump’s pledge is that even if we do bring back manufacturing, it won’t produce jobs because of automation.
Atkinson: That’s not the case. Most of the pundits who make these claims don’t bother to look at the federal government productivity data. ITIF conducted an analysis and we found that if the U.S. could eliminate the manufacturing trade deficit, and if they did it by raising productivity by 16 percent, which we thought was a reasonable number, we’d still create at least 1.5 million manufacturing jobs. And by the way, these jobs have multiplier effects that are much larger than those from creating a job in a restaurant, for example.
Judis: But isn’t the argument that manufacturing jobs have been disappearing because of automation?
Atkinson: The sad reality is that most economists and pundits continually parrot the view that manufacturing jobs have been disappearing because of automation. But that’s just not true. From 2000 to 2011, almost 35 percent of U.S. manufacturing jobs were lost. Our estimate is that at least half of the loss of jobs was due to trade and lack of U.S. competitiveness. That’s also similar to the Economic Policy Institute’s and MIT economist David Autor’s estimates.
Conventional economists argue that because manufacturing output grew almost as fast as GDP, job loss couldn’t be because we were simply producing less. It had to be because of increases in productivity from automation. There are, however, two big problems with this argument.
The first is that the difference between the rate of productivity growth in manufacturing and in the rest of the economy in the 1990s, when we lost just three percent of our manufacturing jobs, was actually larger than the gap in the 2000s when we lost one-third of jobs. So something more than productivity growth had to be responsible for the job loss in the 2000s.
So what could have been going on here? In the late 1990s, the government switched how it measured the output of computers and semiconductors. The government used a new methodology that completely distorted the manufacturing output and productivity numbers. When you bought a new laptop in 2004, and you bought another one in 2007 and it was five times faster and more powerful, the government recorded the company as making five times as many computers. This overstatement in output was so massive that it meant that 110 percent of U.S. manufacturing output growth from 2006 to 2013 came from that one sector, computers and electronics.
The output in the rest of the manufacturing sector, which accounts for about 92 percent of overall manufacturing, actually fell during that period. They were making less here because imports went up. That’s a big reason manufacturing jobs fell as fast as they did. But the Washington trade establishment is so convinced that trade cannot be the problem that they can’t even be bothered to look at the data carefully. They worry that if you acknowledge that manufacturing jobs declined because of trade, the next step is mindless protectionism.
Judis: Let’s go back to the issue of trade. Trump has charged we are losing jobs because of “bad trade deals.” Were jobs lost because other countries made things better and cheaper than we did, or because of unfair trade practices by other countries?
Atkinson: So I believe there were several causes of our trade deficit. One was systematic unfair practices by our trading partners and in particular by China. In the 2000s, that included currency manipulation, forced technology transfer and intellectual property theft, among other unfair practices. The Chinese government was saying that if you don’t make it here, you can’t access their market. And China was simply refusing to import as much as they should have from the U.S.
I’ll readily acknowledge that if, after China had come into the World Trade Organization, it had lived up to its obligations, there would have still been disruption in the U.S. manufacturing sector. We would have lost jobs in industries like textiles, furniture and metal assembly. But the losses wouldn’t have been as dramatic and as fast, and we would have made up for them with greater exports of products like machine tools, scientific instruments, jet airplanes, computers and semiconductors. We would have been creating many more offsetting manufacturing jobs if China hadn’t been cheating.
The second reason for our trade deficit has been the low rate of investment by companies in America. If you add up the amount of total capital assets—all lathes and the machine stampers and other machinery and capital equipment that manufacturers in the United States have—the amount went up between 20 and 50 percent every decade from the 1950s to the 1990s. But in the 2000s, it went up a measly three percent. That had never happened in America before.
So why the stagnation? First, companies were responding to the China shock, in part by moving production to China; but Wall Street also pushed them to do things more short-term where investing in the future was not rewarded. Why invest in research and development, workforce training or new machines and equipment if Wall Street punishes you? So you had a vicious circle.
You hear advocates for manufacturing saying that the U.S. costs are too high and that is why we can’t compete. But if you look at the countries that are doing well in manufacturing, such as Germany, German manufacturing wages are about 45 percent above ours, and yet they run a trade surplus with China. And as you know, many of their manufacturers are unionized.
Judis: How do the Germans succeed where we don’t?
Atkinson: A key difference between us and the Germans is that they have a world class apprenticeship program, and they work very hard to align their technology modernization with workforce skill upgrading. They are competing on the high road.
Austria follows a similar industrial strategy. I once talked to a group of Austrian manufacturing company CEOs and I asked them why they didn’t move as much of their production to China, and they said, “Well it’s not like we don’t move anything to China, but our first response rather than going to our accountants is to go to our engineers and ask them what do you need to do to be able to re-engineer the work so that we can beat the China price. Sometimes we can do it, and sometimes we can’t.” So they are more willing to take a more longer-term approach to try to cut the costs through skill and technology upgrading.
The Germans and Austrians also have a system where they have public-private R&D partnerships for their companies. And it’s not just about automation. It’s also about making the kind of products that China or India can’t make because they are too advanced and complex. The Germans, for example, have a system of over 60 Fraunhofer Institutes that bring together companies to cooperate in particular technology areas, such as industrial engineering, chemical technology, lasers, optics and even advanced wood products. They use science and engineering to make things that low-wage countries can’t make and can’t easily copy.
We’ve only just begun to do that. President Obama put in place the National Network of Manufacturing Innovation, which ITIF first proposed and which ended up getting bipartisan support in Congress. But we’ve started down that path a day late and a dollar short. The Germans have been doing it for forty years and today invest significantly more in their systems than we do. So do many other competitor nations, including the U.K.
Judis: I have also heard the argument that our corporate tax rates are too high in comparison to other countries.
Atkinson: I think that taxes do matter, although lower rates are not a panacea. Look at other countries, most have very generous incentives for companies to invest in research and development, new equipment and upgrading worker skills.
One recent study published by the National Bureau of Economic Research looked at the effective tax rate, not the statutory tax rate, for manufacturers among thirteen or so of our major competitors, and it found that the U.S. was the second highest of the group. Only Japan was higher, but then Japan ended up lowering its rate.
So I do think there is a case to be made for lowering the effective business tax rate, in part by expanding of incentives for investment. For instance, when [Bill] Clinton was in office, we had among the most generous research and development tax credit in the world, now we’re 27th.
Judis: Let’s talk now about what Trump has said and done to date. First, let me ask what you think about Trump jawboning Carrier, Ford, Nabisco and Toyota not to move their factories out of the United States. Critics have called it good politics, but bad economics.
Atkinson: I mentioned the story about the Austrian CEOs. I asked them about why there were slower to move production to China. I thought they were going to say they were worried about the Chinese stealing their intellectual property or something like that. Their answer was interesting. They said that if we don’t at least try to make it in Austria, if we don’t at least bring our engineers in, we will be socially shunned. That’s the term they used.
That is really interesting, and the Germans are like that too. There is a nationalistic sentiment there. The Austrian and German people at least want you to give it the old college try, but here there is just no expectation of that. If you are a CEO, you are beholden to only your shareholders. And the Washington consensus holds that you shouldn’t be beholden to anyone else. And so the generous way to think about what Trump is doing is that he is providing a little bit of social pressure. He is telling companies that he wants them to think twice. He wants them to see if their engineers can find a way to make that product line in Indiana. That’s the good side of what Trump is doing.
Now I see two potential downsides. One is that if that’s all he wants to do, it’s not enough. As I said, the Germans and the Austrians have extremely well developed manufacturing support systems in place. We have a program run out of the National Institute of Standards and Technology (NIST) called the Manufacturing Extension Partnership (MEP). It’s kind of like the old Agricultural Extension Partnership. It’s a very good program that helps small manufacturers with adopting new technology and training their workers, but the Canadians spend ten times more than we do per GDP on these kind of ventures. The Germans spend twenty times, and the Japanese forty times. So all the jawboning in the world is not going to make up for gaps like this in public investment.
Unfortunately, a first look at the proposed Trump budget would suggest this will get worse, not better. For example, his budget would eliminate the already underfunded MEP program while cutting funding for an array of R&D programs whose discoveries help U.S. manufacturers. For a candidate who campaigned on restoring U.S. manufacturing jobs, his budget will do the exact opposite.
The other thing that worries me about Trump’s approach is that it’s one thing to say we want to pressure you to invest in America, but if it crosses the line into moving away from our rule of law, that’s worrisome. For example, if Trump threatens to have the federal government retaliate against companies that don’t do what he wants, that would be worrisome. But there’s no evidence he’s done that or plans to do that.
Judis: Trump has spoken repeatedly during his campaign and presidency about our making “bad trade deals” with Mexico and China and promising to revise them. Is Trump on the right track here?
Atkinson: There is a good side and a bad side to this, too. We finally have a President who takes the trade deficit really seriously and who takes foreign mercantilism really seriously. I frankly don’t think the Bush administration or the Obama administration took this as seriously as it deserves. They were really in the market-opening mode rather than in the trade-enforcing mode. So that’s a useful rebalancing that Trump can bring to the table.
However, I do have some worries about his trade policy. I listened to [Director of the White House’s National Trade Council] Peter Navarro’s speech at the National Association of Business Economists. He implied that the U.S. should be in balance with all nations and that if we’re not in balance with one country, that requires action. That would be a huge mistake. The goal should not be to balance trade with every country. The goal should be to balance or be close to balance with the whole world, so you can be running a trade surplus with Austria and a trade deficit with the Mexico. That’s the way it ought to be.
I don’t agree that NAFTA [North American Free Trade Agreement] is a bad deal. Mexico is fundamentally not a mercantilist nation. They may be doing a few things here and there, just as Canada does, but Canada and Mexico are not mercantilist nations. Why then alienate them? Why fight a fight that doesn’t need to be fought?
The fight we have to fight is against nations that are systemic mercantilists. And China is number one on the list. Ninety percent of our effort on trade enforcement should be against China . They are the big elephant in the room. Their dominant logic regarding globalization is mercantilism. So I worry that Trump is picking fights with nations who fundamentally should be our allies. Alienating nations like Mexico and Germany is to no avail in my view.
There is one other thing I worry that the Trump administration does not understand. In his speech Peter Navarro said we are going to get trade agreements to boost things like wine exports to help Sonoma County, and apricot exports, etc., but these don’t fundamentally matter to our nation’s economic future. This is fighting the past war. What really matters is how we are faring globally in the industries of the future like: aviation, aerospace, biotechnology, semiconductors, software and the internet.
If we lose significant global market share in these industries, America’s future is bleak. We could end up like New Zealand. New Zealand doesn’t run a big trade deficit because they export large quantities of wool and mutton. Is that really what we want? So I worry that Trump is not going to be strategic. It’s just not important to defend our furniture or chicken industries. It is important to make sure that we are well positioned for the industries of the next forty years. We don’t want the Chinese, and the Brazilians, and the Russians and the Indians to take us to the cleaners on those industries.
Judis: Trump has repeatedly talked about reviving the steel industry, and Trump’s Secretary of Commerce Wilbur Ross is known for having revived a steel company.
Atkinson: That’s a big concern. I worry that their ideal company is a steel company that has no operations outside the United States. But steel, aluminum, and other heavy industries are not the industries of tomorrow. I am not saying throw them overboard. In China, there is government-enabled overcapacity in a lot of these industries that has led to dumping in the U.S. market. So yes, it’s worth taking action, but the number of aluminum smelter workers in the U.S. is not very big. If the administration is really going to make America great economically it needs to focus its attention on the advanced industries of tomorrow.
I am also worried that they don’t understand the relationship between the health of multinationals and of the U.S. domestic economy. The idea that the U.S. could be successful by becoming a pre-Roosevelt national economy makes no sense. You can’t do that anymore because the scale of advanced industries need to be successful is so big. Aviation, aerospace, the internet: all have to serve global markets if you want to win globally.
So I am worried that they look at those industries and they say they are Benedict Arnold firms because they moved some operations to China. To me that’s an emotional response. The right response is how are we going to structure trade and domestic competitiveness policy to advance American interests and not get into a tiff because someone may think some of these companies are being patriotic.
Judis: Taxes. What about this border tax on imports and not on exports that the House leadership has backed and that Trump seems to have endorsed?
Atkinson: Most countries have a value-added tax that they impose, kind of like a sales tax, and it is structured so something coming into the country pays the tax, and something going out of the country is exempt from the tax. When the U.S. tried to do something like it before—it was called FSC ETI [Foreign Sales Corporation Extra-Territorial Income]—and gave exports a tax break, and the WTO said you can’t do it. It was a really ridiculous decision given that other nations do the essentially the same thing, but that’s what they said.
So what the House Republicans have done is propose a border adjustable tax to raise a lot of money so that the proposed cut to the corporate tax rate is revenue neutral. On the one hand, this is a response to the inequity around the value added tax other countries have, and it raises money to make a lower rate or a better research and development credit affordable.
But it may be difficult to implement. It might be better to just implement a straight-up value-added-tax, but many Republicans don’t want that, because they think it will just enable bigger government. So if they do put in place a border tax, it would be important to phase it in over time so that companies can adjust their supply chains over time.
Judis: We know that Trump wants to eliminate the Manufacturing Extension Partnership. Have you heard whether the Trump administration is going to budget the Obama administration’s other effort at industrial policy, the National Network of Manufacturing Innovation (now called Manufacturing USA)?
Atkinson: I have not heard details on this, but looking at his Heritage Foundation-inspired budget cutting plan, it would lead one to think that is a distinct possibility. However, some of the Manufacturing USA institutes are funded with DOD money so perhaps these ones might not be on the chopping block. But overall it appears that the Trump administration wants to gut needed public investment and support manufacturing instead with high tariffs and other forms of protectionism. In other words, they want to cut funding for programs and policies to help the U.S. economy compete better globally, while at the same time implementing protectionist policies.
It is one thing to punish some countries with the idea that you are doing it as a negotiating tactic so they stand down their mercantilist policies and once it’s gone you are back to normal. It’s another thing to try to emulate them and become protectionist. Likewise, it’s one thing to try to cut programs that aren’t working well or that go purely to domestic consumption. It’s quite another thing to gut public investment. This reversion to a pre-New Deal Republican party—a part of small government mercantilism—is an anachronism that will only accelerate U.S. economic competitiveness decline.