Now that secular stagnation has become a catchword in financial circles, everyone seems to be weighing in on it. In a series of recent tweets, Marc Andreessen, a Silicon Valley honcho, has confidently declared that secular stagnation poses no risk to the US economy. In support of his argument, Andreessen cites a convoluted report from the Bank of Italy and quotes Larry Summers, who stated on CNBC that secular stagnation is more a problem for Japan and Europe than the US. The mistake in Andreessen’s logic, however, and it’s a mistake that many commentators make, is to confuse two definitions of secular stagnation. The first is a supply-side definition, and the second is a demand-side definition.
In a gist, what the supply-side definition says is that advanced countries will fall behind in productivity improvements (making more with less) and therefore the potential growth rate of the economy will slow. The reason why Andreessen isn’t swayed by this argument is because, as a creature of tech, he is sanguine about future innovation. And on this point he’s probably correct. This Malthusian pessimism, after all, has been dogging us for centuries, and it’s always wrong.
But this doesn’t make secular stagnation benign, because it doesn’t consider the demand-side definition of the term, which is the one I care about it. The demand-side version says that, because of deficient spending, our economies will no longer be able to put all willing workers to work. Prolonged weak demand, in other words, will keep downward pressure on GDP, prices, and jobs. More technically, we can say that the interest rate necessary for full employment (what I refer to as “the ideal interest rate”) is negative when the government’s budget is balanced.
Readers of my book will notice that this definition is somewhat broader than the one employed in Chapter 6, which stated that secular stagnation occurs when the ideal interest rate is negative in the absence of government deficit spending. At the time of writing, I was too narrowly focussed on government deficits, perhaps swayed by the fact that it has been decades since we last saw a fiscal surplus.
This oversight in no way detracts from my core argument, but for the sake of clarity, let’s take a moment to explain it anyway. Remember that the ideal interest rate is the rate that balances saving and borrowing at full employment, which is necessary for the economy to remain fully employed. This differs from the “actual interest rate,” which is the rate manipulated by the central bank. The ideal rate is determined by two factors: the supply for savings and the demand for borrowing. How much does the economy want to save at a given interest rate? How much does it want to borrow? The private sector is the main determinant of these preferences, but the government can also influence them by altering its deficit or surplus. For instance, if the actual interest rate stays the same, and the government increases its deficit from -1% of GDP to -5%, the ideal interest will rise. In essence, what the government is saying here is this: “I am now willing to borrow more at a given interest rate.” Assuming the private sector doesn’t then offset this change in demand, the additional borrowing would lift the overall economy.
Contrarily, if the government suddenly ran a surplus of 10% of GDP, the economy would crumble. In this case the ideal interest rate would certainly be negative, but does that mean the economy is suffering from secular stagnation? Far from it. What this economy is suffering from is deranged fiscal policy. It’s very likely that the ideal rate would no longer be negative if the government reduced its surplus to 0%.
This is why, when defining secular stagnation, we need think of an economy where the government is neither adding to borrowing demand (fiscal deficits) nor to the supply of savings (fiscal surplus).
Okay, now let’s go back to our core argument. How do we decide whether an economy is at risk from secular stagnation? In my book, I focus on both sides of the ideal interest rate: the demand for borrowing and the supply of savings. On the savings side, I argue that because of corporate hoarding and saving for retirement, there will always be a plentiful supply of savings at a 0% interest rate in a full employment economy. These excess savings will only increase as our economies become richer and more unequal, since rich nations tend to save more than poor ones, and, within an individual economy, the rich tend to save more than the poor. So we would expect an advanced economy suffering from acute income inequality to be especially prone to high savings rates.
But on the borrowing side, there is no guarantee that these savings will be absorbed by a full employment economy at a 0% interest rate. The main reason people borrow is to invest. For developing countries, capital-intensive catch-up productivity investments can often be enough to absorb an economy’s savings. But in advanced nations with already high productivity, the major driver of investment demand is the growth in its potential workforce. In other words, it’s demographics. New workers need new houses and new offices, new cars and new roads. And as the rate of workforce growth slows, so does the demand for borrowing at a given interest rate.
How bad is this demographic trend? Here is a chart that predicts Japanese potential workforce growth based on census data. The 1950 line, for example, uses 1950 census data to calculate the annual growth rate in the potential work force as the years went on. The source is the United Nations.
Notice how the large drop in workforce growth coincides almost perfectly with the start of Japan’s Lost Decades.
Now here is the same chart for the US.
While these two charts alone don’t definitively prove that secular stagnation will afflict the US, what’s clear is that we shouldn’t dismiss the risk out of hand. Larry Summers, in other words, may want to reassess his recent optimism.