(Analysis) American labor productivity slowed noticeably in the 1970s. When it began to pick up in the ’90s, many observers concluded that the slowdown was a temporary incident due to inflation or supply shocks. Unfortunately, labor productivity again slowed in the early 2000s. This is puzzling to many, and the analysis here is offered as a (partial) explanation by way of economic “openness”.
Labor productivity is at least a monotonic function of capital per worker – the more machines and infrastructure available to labor, the higher should be labor’s productivity. This is straightforward and should be uncontroversial. What comes next is not uncontroversial, though. Is it possible that increasing openness in our economy – measured by trade and direct investment flows overseas – is leading to a slowdown of American labor productivity? Data for openness in the US economy suggest this very possibility.
Looking at a relative measure for trade on a balance-of-payments basis – the sum of exports and imports as a share of GDP – in the first chart (above left) shows that trade is increasing as a share of our GDP. This is a measure used often by economists to capture the openness to trade by an economy. It is interesting that this measure is negatively correlated with an index for labor productivity after differencing. The correlation is -0.23 over the years 1960-2012. While correlation does not imply causality, it does not rule it out, either.
Another measure for our economy’s openness is the outflow of direct investment to other countries. This is meant to capture capital that is seeking a higher return external to the US, meaning that it is not being invested domestically. Similarly, this measure is negatively correlated with an index of labor productivity after differencing. The correlation is -0.21 over the years 1982-2012.*
Both these measures of our economy’s openness – trade and direct investment overseas – are negatively related to labor productivity. What is going on here? Isn’t trade supposed to improve productivity through international competition?
Well, if we assume that there is something going on, whereby openness is affecting aggregate labor productivity in a negative way, there must be a channel for this to occur. Two possible explanations include less capital per worker in the US than there otherwise would be in the absence of openness (due to leakage through overseas investment), and shifting of overall domestic production to less capitally intensive production. The first explanation is straightforward; the second is more difficult to understand, though, since the US is generally assumed to hold a comparative advantage in capitally intensive goods.
One possible hypothesis to explain why production is shifting to less capitally intensive technologies is that there is a growing split of the US economy into two diverging economies: one that is capitally intensive with high labor productivity, and the other, its opposite, which is growing as an overall share of the American economy with measures of openness. Based on the data shown, this could explain sluggish labor productivity, but much more thorough analysis is required before someone could conclude either way with any degree of certainty. Another possible hypothesis for a shift to less capitally intensive technologies – admittedly a long shot – is that capital re-switching or reversing has occurred, where (relative) scarcity of capital does not determine a higher price of capital. In other words, scarcity isn’t determining (efficiently?) the ratio of capital to labor inputs per neoclassical economic assumptions. This oddity of capital’s pricing was the subject of the so-called Cambridge controversy in the 1950s and ’60s.
One thing is sure, though: openness in the US economy to trade and overseas direct investment is not unambiguously positive in its effect for the economy. If gains from trade are so skewed that the overwhelming majority is less well off from trade, then there is a problem with trade-driven economic growth. This contradicts much simplified trade theory, but would seemingly explain much of our dilemma during the past decades – if trade is negatively affecting American labor productivity.
*Data are available from the source only from 1982.