December 05, 2019
(This piece originally appeared on my Patreon page.)
Economists are continually developing new statistical techniques, at least some of which are useful for analyzing data in ways that allow us to learn new things about the world. While developing these new techniques can often be complicated, there are many simple things about the world that economists tend to overlook.
The most important example here is the housing bubble in the last decade. It didn’t require any complicated statistical techniques to recognize that house prices had sharply diverged from their long-term pattern, with no plausible explanation in the fundamentals of the housing market.
It also didn’t require sophisticated statistical analysis to see the housing market was driving the economy. At its peak in 2005 residential construction accounted for 6.8 percent of GDP. This compares to a long-run average that is close to 4.0 percent. Consumption was also booming, as people spent based on the bubble generated equity in their homes, pushing the savings rate to a record low.
The existence of the bubble and the fact that it was driving the economy could both be easily determined from regularly published government data, yet the vast majority of economists were surprised when the bubble burst and it gave us the Great Recession. This history should lead us to ask what other simple things economists are missing.
For this holiday season, I will give three big items that are apparently too simple for economists to understand.
1)Profit shares have not increased much — While there has been some redistribution in before-tax income shares from labor to capital, it at most explains a small portion of the upward redistribution of the last four decades. Furthermore, shares have been shifting back towards labor in the last four years.
2) Returns to shareholders have been low by historical standards — It is often asserted that is an era of shareholder capitalism in which companies are being run to maximize returns to shareholders. In fact, returns to shareholders have been considerably lower on average than they were in the long Golden Age from 1947 to 1973.
3) Patent and copyright rents are equivalent to government debt as a future burden – The burden that we are placing on our children through the debt of the government is a frequent theme in economic reporting. However, we impose a far larger burden with government-granted patent and copyright monopolies, although this literally never gets any attention in the media.
To be clear, none of these points are contestable. All three can all be shown with widely available data and/or basic economic logic. The fact that they are not widely recognized by people in policy debates reflects the laziness of economists and people who write about economic policy.
It is common to see discussions where it is assumed that there has been a large shift from wages to profits, and then a lot of head-scratching about why this occurred. In fact, the shift from wages to profits has been relatively modest and all of it occurred after 2000, after the bulk of the upward redistribution of income had already taken place.
If we just compare end points, the labor share of net domestic product was 64.0 percent in 2019, a reduction of 1.6 percentage points from its 65.6 percent share in 1979, before the upward redistribution began. If, as a counter-factual, we assume that the labor share was still at its 1979 level it would mean that wages would be 2.5 percent higher than they are now. That is not a trivial effect, but it only explains a relatively small portion of the upward redistribution over the last four decades.
It is also worth noting the timing of this shift in shares. There was no change in shares from 1979 to 2000, the point at which most of the upward redistribution to the richest one percent had already taken place. The shift begins in the recovery from the 2001 recession.
This was the period of the housing bubble. The reason why this matters is that banks and other financial institutions were recording large profits on the issuance of mortgages that subsequently went bad, leading to large losses in the years 2008-09. This means that a substantial portion of the profits that were being booked in the years prior to the Great Recession were not real profits.
It would be as though companies reported profits based on huge sales to a country that didn’t exist. Such reporting would make profits look good when the sales were being booked, but then would produce large losses when the payments for the sales did not materialize, since the buyer did not exist. It’s not clear that when the financial industry books phony profits it means there was a redistribution from labor to capital.
There clearly was a redistribution from labor to capital in the weak labor market following the Great Recession. Workers did not have enough bargaining power to capture any of the gains from productivity growth in those years. That has been partially reversed in the last four years as the labor share of net domestic income has risen by 2.4 percentage points. This still leaves some room for further increases to make up for the drop in labor share from the Great Recession, but it does look as though the labor market is operating as we would expect.
Returns to Shareholders Lag in the Period of Shareholder Capitalism
It is common for people writing on economics, including economists, to say that companies have been focused on returns to shareholders in the last four decades in a way that was not previously true. The biggest problem with this story is that returns to shareholders have actually been relatively low in the last two decades.
If we take the average real rate of return over the last two decades, it has been 3.9 percent. That compares to rates of more than 8.0 percent in the fifties and sixties. Even this 3.9 percent return required a big helping hand from the government in the form a reduction in the corporate income tax rate from 35 percent to 21 percent.
The figure for the last two decades is somewhat distorted by the fact that we were reaching the peak of the stock bubble in the late 1990s, but the story is little changed if we adjust for this fact. If we take the average real return from July of 1997, when the price to earnings ratio was roughly the same as it is now, it is still just 5.7 percent, well below the Golden Age average when companies were supposedly not being run to maximize shareholder value.
It is striking that this drop in stock returns is so little noticed and basically does not feature at all in discussions of the economy. Back in the late 1990s, it was nearly universally accepted in public debates that stocks would provide a 7.0 percent real return on average in public debates.
This was most evident in debates on Social Security, where both conservatives and liberals assumed that the stock market would provide 7.0 percent real returns. Conservatives, like Martin Feldstein, made this assumption as part of their privatization plans. Liberal economists made the same assumption in plans put forward by the Clinton administration and others to shore up the Social Security trust fund by putting a portion of it in the stock market. The Congressional Budget Office even adopted the 7.0 percent real stock returns assumption in its analysis of various Social Security reform proposals that called for putting funds in the stock market.
Given the past history on stock returns and the widely held view that returns would continue to average close to 7.0 percent over the long-term, the actual performance of stock returns over the last two decades looks pretty disappointing from shareholders’ perspective. It certainly does not look like corporations are being run for their benefit, or if so, top executives are doing a poor job.
One of the obvious factors depressing returns has been the extraordinary run up in price to earnings ratios. A high price to earnings ratio (PE) effectively means that shareholders have to pay a lot of money for a dollar in corporate profits. When PEs were lower, in the 1950s and 1960s, dividends yields were in the range of 3.0 -5.0 percent. In the recent years they have been hovering near 2.0 percent. When the PE is over 30, as is now the case, paying out a dividend of even 3.0 percent would essentially mean paying out all the company’s profits as dividends. Clearly that cannot happen, or at least not on a sustained basis.
While shareholders have not done well by historical standards in recent decades, CEO pay has soared, with the ratio of the pay of CEOs to ordinary workers going from 20 or 30 to 1 in the 1960s and 1970s, to 200 or 300 to 1 at present. There is a story that could reconcile soaring CEO pay with historically low stock returns.
Corporations have increasingly turned to share buybacks as an alternative to dividends for paying out money to shareholders. The process of buying back shares would drive up share prices. Part of this is almost definitional, with fewer shares outstanding, the price per share should go up. If buybacks push up share prices enough to raise the price to earnings ratio, then in principle other investors should sell stock to bring the PE back to its prior level. But if this doesn’t happen, then buybacks could increase PEs.
That would of course imply huge irrationality in the stock market, but anyone who lived through the 1990s stock bubble and the housing bubble in the last decade knows that large investors can be exceedingly irrational for long periods of time. Anyhow, if share buybacks do raise PEs there would be a clear story whereby CEOs could drive up their own pay, which typically is largely in stock options, to the detriment of future shareholders, which would explain both soaring CEO pay and declining returns to shareholders.
Whether this story of share buybacks raising PE is accurate would require some serious research (I’d welcome references, if anyone has them), but what is beyond dispute is that the last two decades have provided shareholders with relatively low returns. That seems hard to reconcile with the often repeated story about this being a period of shareholder capitalism.
Patents and Copyright Monopolies Are Implicit Government Debt
There is a whole industry dedicated to highlighting the size and growth of the government debt, largely funded by the late private equity billionaire Peter Peterson. The leading news outlets feel a need to regularly turn to the Peterson funded outfits to give us updates on the size of the debt.
When presenting the horror story of a $20 trillion debt and the burden it will impose on our children, there is never any mention of the burden created by patent and copyright monopolies. This is an inexcusable inconsistency.
Patent and copyright monopolies are mechanisms that the government uses to pay for services that are alternatives to direct spending. For example, instead of granting drug companies patent monopolies and software developers copyright monopolies, the government could just pay directly for the research and creative work that was the basis for these monopolies. There are arguments as to why these monopolies might be better mechanisms than direct funding, but these arguments don’t change the fact they are mechanisms the government uses for paying for services.
While we keep careful accounting of the direct spending, we pretend the implicit spending by granting patent and copyright monopolies does not exist. This makes zero sense, especially given the size of the rents being created by these monopolies.
In the case of prescription drugs alone, we will spend close to $400 billion (1.8 percent of GDP) this year above the free market price, due to patent protections and other monopolies granted by the federal government. This is considerably more than the $330 billion in interest that the Congressional Budget Office projected we would spend on the $16.6 trillion in publicly held debt in 2019.
And this figure is just a fraction of the total rents from patent and copyright monopolies, which would include most of the payments for medical equipment, computer software and hardware, and recorded music and video material. Since these payments dwarf the size of interest payments on the debt, it is difficult to understand how anyone concerned about the burdens the government was creating could ignore patents and copyrights, while harping on interest on the debt.
As I have often argued there are good reasons, especially in the case of prescription drugs, for thinking that direct funding would be a more efficient mechanism than patent monopolies. In the case of prescription drugs, direct funding would mean that all findings would be immediately available to all researchers worldwide. If drugs were sold at free market prices, it would no longer be a struggle to find ways to pay for them. And, we would take away the incentive to push drugs in contexts where they are not appropriate, as happened with the opioid crisis. (See Rigged, chapter 5, for a fuller discussion [it’s free.])
While the relative merits of patent/copyright monopolies and direct funding can be debated, the logical point, that these monopolies are an implicit form of government debt, cannot be. It shows the incredibly low quality of economic debate that this fact is not widely recognized.
The Prospect for Simple Facts and Logic Entering Economic Debate in the Next Decade
The three issues noted here are already pretty huge in terms of our understanding of the economy. The people who write in a wide range of areas should be aware of them, but with few exceptions, they are not.
Unfortunately, that situation is not likely to change any time soon for a simple economic reason, there is no incentive for people who write on economic issues to give these points serious attention. They can continue to draw paychecks and get grants for doing what they are doing. Why should they spend time addressing facts and logic that require they think differently about the world?
As has been noted many times, there is no real consequence to economists and people writing about the economy for being wrong. A custodian who doesn’t clean the toilet gets fired, but an economist who missed the housing bubble whose collapse led to the Great Recession gets the “who could have known?” amnesty.
Given this structure of incentives, we should assume that economists and others who write on economics will continue to ignore some of the most basic facts about the economy. That is what economics tells us.
 Since income is supposed to be matched by output in the GDP accounts, the corresponding phony entry on the output side would be the loans that subsequently went bad. These loans were counted as a service when they were issued. Arguably, this was not accurate accounting.
 This rise in labor share appears in the net domestic income calculation, but not in the net domestic product figure. The reason is that there has been a sharp drop in the size of the statistical discrepancy over the last four years, as output side GDP now exceeds the income side measure. It is common to assume that the true figure lies somewhere in the middle, which would mean the increase in labor share is likely less the 2.4 percentage points calculated on the income side.