The quixotic quest to improve productivity growth
NRPLUS MEMBER ARTICLE
he United States has been worrying about productivity for the whole of my life, and for the whole of my life, we have made almost no progress in figuring out where productivity growth really comes from or how to cultivate it.
There is a very powerful political incentive to do so: By one conventional measure — and apply here the usual caveats — American men’s median earnings peaked in 1973 and were in real-dollar terms about 4.4 percent lower by 2012. The growth in household income in my generation has come from the increasing participation of women in the workforce and the rise in women’s real wages, which peaked around 2007 and have not yet quite recovered from the Great Recession. Americans 50 and younger have had a very different experience of intergenerational prosperity than their parents did. They are frustrated.
The postwar economic boom in the United States was based on a combination of unusual factors that cannot be repeated and that we mostly would not want to repeat even if we could: Reducing Europe, the United Kingdom, and Japan to smoking ruins — some of them irradiated, others languishing under brutal occupation by a federation of police states affiliated with the Union of Soviet Socialist Republics — is not exactly a promising policy for the 21st century.
But why did the postwar boom end so suddenly? What in Hell happened in 1973? From 1948 to 1969, nonfarm labor-productivity growth averaged 2.5 percent. That looks like a very modest number, but it implies a radical improvement in standards of living from one generation to the next. From 1969 to 1973, it slowed a bit, to about 2 percent. After that, it came almost to a halt, averaging 0.5 percent from 1973 to 1979. As economist Alicia Munnell put it in a paper written 30 years ago — like I said: We’ve been worrying about this forever — the 2.5 percent productivity growth of the postwar era meant a doubling in living standards every 28 years, whereas at 0.5 percent those same 28 years would see only a 15 percent growth in living standards.
Productivity growth (meaning still nonfarm labor-productivity growth) increased toward the turn of the century, driven in part by advances in technology, especially the proliferation of personal computers and then the growth of the Internet. In 1998, productivity growth was at a very encouraging 3.8 percent. In 2002, it was at 4.4 percent. (Labor productivity tends to spike during and immediately after recessions; this could be a kind of bookkeeping artifact or the result of businesses’ firing their least-productive workers first. The explanation preferred by some economists at the National Bureau of Economic Research is terror.) But productivity growth averaged less than 2 percent from 1993 to 1997, which were pretty good years, and after the turn of the century it declined steadily to 1.2 percent in 2006.
Compounding the problem is the fact that even the modest productivity growth we have seen has not been accompanied by a parallel increase in real wages. This “decoupling” of wages and productivity is not unique to the American economy but also is present in the economies of most of the OECD countries, including in the European welfare states that American progressives say they admire. In Germany and Ireland, for example, productivity increased at three times the rate of real wages from 2000 to 2016; in Belgium and Austria, productivity has increased only twice as fast as real wages. In some of the less robust European economies, such as Hungary, Portugal, and Greece, real wages went down even as labor productivity increased.
Where does productivity growth come from? That is an easy question to answer — retrospectively. It is a very difficult question to answer with an eye to the future.
Most modern industrial economies will have gone through a period of rapid productivity growth at some point in their histories. Most often, that is the result of an agrarian economy’s turning into an industrial one — a worker in a 19th-century factory produced a lot more value in an hour’s work than did a worker on a 19th-century farm. (There is a reason economists specify nonfarm labor.) Most of the economic “miracles” of centrally planned economies were roughly that — one of the things about having a brutal totalitarian state is that you can drive the peasants off of the farms and into the factories, if you want to. Other countries have seen dramatic growth in prosperity driven by new trade opportunities, which increase the value of existing production.
There are other sources of productivity growth, including education and training that make labor more valuable, infrastructure projects that enable more efficient use of land and other resources, and, probably most important in the American context, innovation and the investment to implement it. But innovation is a will-o’-the-wisp — it cannot be captured and put in a cage to be trotted out when needed, like a draft horse. The technological innovations that drove the prosperity of the 1990s came in no small part from research undertaken as early as the 1940s and developed in the 1950s and 1960s, when the first computers were connected to what would become the Internet. Packet-switching was developed by RAND and the Defense Department back in the 1960s. The “frequency hopping” technology supporting our wireless devices was on the mind of Marconi in the 19th century and was being used in military applications by the time of the Great War. But all of that was a long way from the iPhone, or even attaching a .doc to an email.
In the United States, we spend significantly more per student on K–12 education today than in past generations, we send a much greater share of the population to college than we did before, and we spend more per student on higher education than any other country save Luxembourg, where the professors must be living large, indeed. We do not have much to show for that when it comes to productivity growth. We have a gazillion worker-training programs implemented at virtually every level of government, to very little effect. The welfare state created through the New Deal and Great Society has not improved productivity in the way many of the proponents of those programs thought (or at least said) it would — but, then, the Reaganite program of tax cuts and deregulation has not brought back that 2.5 percent productivity growth of the postwar years, either, at least on a long-term basis. We have reached into a lot of hats — still no rabbit.
As noted earlier, the political incentives for figuring this out are very strong, and there is not really any sensible incentive for not improving productivity growth, unless you count the anti-growth (and anti-human) agenda of the radical environmentalists. This is the sort of thing that inspires conspiracy theories and myths — you know, that secret automobile engine that runs off of seawater that Big Oil has been keeping under wraps all these years. The more reasonable conclusion is that we do not know how to improve productivity in any direct and reliable way, and that this may in fact be not knowable at all. There is not any formula for better productivity, stronger growth, higher wages, or prosperity that is more robust, more widely shared, and more confidence-inspiring. The people who tell you otherwise are selling something, usually themselves.
There is no magic bullet, but there are some things that have stood the test of time: the rule of law, property rights, an open economy, prudent government, democratic accountability, an enlightened understanding of public goods, the skillful balancing of adaptability, stability, and predictability in public policy. The great tragedy of the moment is that the populist movements inspired in no small part by disappointment with economic conditions (and social conditions intertwined with those economic conditions) are fundamentally hostile to the most reliable aspects of the system we have inherited. They seek to elevate strongman executives over the slow work of legislators (the effective abnegation of Congress has not helped here), factional fervor over consensus, a politics of spoils over property rights, and urgent tribal demands over the rule of law. As a result, we enjoy the benefits of neither predictability nor adaptability — instead, we lurch from enthusiasm to enthusiasm and from outrage to outrage: Should we knock over the statue of this Jefferson (Davis) or that Jefferson (Thomas), or neither, or both? It is all very exciting. Much more exciting than work.
These things are linked: It is not entirely an accident that the postwar productivity boom ended shortly after the riots of the late 1960s. Philadelphia, to take one example, lost a fifth of its population and a quarter of its jobs second half of the 20th century — it has fewer residents today than it did in 1920. In the United States, our centers of productivity are the cities and the energy corridor. What is bad for city life is bad for the economy and bad for the country at large, and the current disorder is bad for city life on both political fronts: The progressives and their instrument, the Democratic Party, are cowed into acting as apologists for rioting and arson, their energies turned to the pursuit of daft policies and demonstrations of tribal allegiance; conservatives and their instrument, the increasingly rural and exurban Republican Party, are confirmed in their contempt for the cities and the people who live there, and so urban problems are at the bottom of their to-do list except when those can be used to inconvenience or embarrass some Democratic political interest.
If we act as if we do not know what is good for us, it is because we don’t. But we also act as if we don’t care to do what’s good for us, because we don’t, preferring instead to service our private rages and petty grievances. There can be no excuse for that, and no forgiving it.