Friday, February 17, 2017

We are the 100%.

Eons ago, I told commenter TravisV that I intended to look at this paper, and I finally have.  The abstract is:
Three mutually uncorrelated economic disturbances that we measure empirically explain 85% of the quarterly variation in real stock market wealth since 1952. A model is employed to interpret these disturbances in terms of three latent primitive shocks. In the short run, shocks that affect the willingness to bear risk independently of macroeconomic fundamentals explain most of the variation in the market. In the long run, the market is profoundly affected by shocks that reallocate the rewards of a given level of production between workers and shareholders. Productivity shocks play a small role in historical stock market fluctuations at all horizons.
This would appear on the surface to push against the notion that we are the 100%.  Over the long term, fluctuations in stock market value come from reallocation between workers and shareholders.

The actual findings of this paper are interesting and useful, but I think we need to be careful about how they are interpreted.  Much like the Mian & Sufi findings about the housing boom, mostly what is going on here is that they have adjusted away almost the entire story, and they are analyzing the small sliver that is left.  They have detrended the data exponentially.  For instance, take a look at this graph from the voxeu article:

Over the long term, we can see here that fluctuations from the trend largely correlate with changes in the share of income to shareholders.

From the article's conclusion:

Technological progress that raises aggregate consumption and benefits both workers and shareholders plays a small role in historical stock market fluctuations at all horizons...
Indeed, without these shocks, today's stock market would be about 10% lower than it was in 1980. The shocks responsible for big historical movements in stock market wealth are not those that raise or lower aggregate rewards, but are instead ones that redistribute a given level of rewards between workers and shareholders.
This seems like misleading interpretation to me.  We are talking about a 10% fluctuation over a period where the trend growth was something like 700% in real terms.

Here is a scatterplot of real capital income and real compensation since WW II.  If you want to know what capital income will be in a given year, an extremely good proxy would be knowing the level of compensation in that year - and vice versa.  The correlation is .97.

So, whatever we might learn from this paper - and there are things to learn from it - it seems very important to keep in mind that this paper is about a 3% portion of the total story.

It seems to me that the quote above should be prefaced with the sentence: Technological progress that raises aggregate consumption and benefits both workers and shareholders explains general growth in stock market values, which is about 97% of the growth in income and wealth.  Shifting factor shares might explain much of the other 3%.

It's a shame that the human psyche is so drawn to battles over relative status.  The story of human history and human advancement is a story of the battle to overcome this mental defect.  We are the 100%.  How much social attention is paid to the 97% of the story versus the 3%?

The largest risk of economic dislocations, like what we have seen over the past couple of decades, isn't the actual shocks themselves.  It is the human tendency to retreat into battles over relative status.  Notice that their measure of the effect of factor shares on stock wealth has declined since the late 1990s.  How's that workin' for ya?  Is there any disagreement that 1968 and 1998 were better for both shareholders and workers than 1978 or 2008?



  1. I could easily flip it around and say "it's a shame that humans can't learn to be equally motivated within massively-redistributive systems." You are making a value judgement that one "DNA-flaw" is worse than the other.

    1. There are many, many more things going on there besides motivation from relative income. The relationship between capital and labor income itself is deep and inscrutable. You're also interpreting this as an inequality issue, wish is a different topic. Much income inequality and much potential redistribution happens within labor income. Capital income and compensation income a very different animals. One is income from delayed consumption and from exposure to risk and loss and the other is income from personal effort. They don't even share the same basis.

    2. One more point. One take-away from this is that capital and labor are overwhelmingly complements. By inserting the redistributive notion you are rhetorically denying that fact, which I am arguing is destined to fail.

    3. Understood. All i'm saying is there are two ways to "solve" battles over relative status: 1) insist people stop caring about status, 2) make greater efforts to equalize status. Neither has the moral high ground in my mind.

    4. I am not arguing against legitimate efforts to equalize status. I am arguing against illegitimate efforts to equalize status. It is highly counterproductive to focus relative status efforts on relationships that are practically mathematically fixed. There are many sources of relative well-being that aren't fixed, complex, complementary relationships. Those would be places to look for equalization. The evidence on the capital/labor divide is overwhelming. Attempts to improve compensation by 20% in 20 years by transferring that income from capital will most likely end up mostly reducing income of both. Framing such as what this paper appears to do gives the impression that relative shares are overwhelmingly important regarding capital income when, in fact, the opposite is the case.

    5. Agreed. But let's also not forget that a big part of the argument is the distribution within labor.

  2. «Technological progress that raises aggregate consumption and benefits both workers and shareholders explains general growth in stock market values, which is about 97% of the growth in income and wealth»

    That is missing the effect of workforce growth: it was 63 million in 1950 and 149 million in 2014. Presumably this means that a factor of 2.2 in growth is due to the "economy" being bigger, not better, so that 97% is wildly overestimated.

    Also median unit wages have not grown much since 1980. Most of the growth in employee income has happened at the top, where there is a significant proportion of rent or profit from "capital".

    1. Good point on the population. Adjusting for the size of the labor force reduces it to 92%. Our shared abundance only accounts for 92% of incomes after the adjustment.

      On the other point, the original article I referenced was referring to corporate ownership and the influence of factor shares on capital incomes. Changes of the composition within factor shares are not pertinent. In fact, that is kind of the point. Trends in compensation distribution have very much to do with changes within the composition of labor and very little to do with the relative income going to capital. If you are saying that some of that shift has come from workers developing "human capital" or from workers receiving deferred compensation in the form of ownership shares, that is still a story about the difference between workers. In any case, as an investor, if I knew that total compensation in 20 years would be no higher than it is today, that would clearly be bad news, not good news, and it would be a tip-off to avoid investing in this economy. That is not the impression I think one would get from the paper I referenced.