I was going to post a bunch of details about how asset prices can be high simply as a function of low required returns, how capital taxation can lower growth rates, how the high growth that came with capitalism transformed the world for all its inhabitants, how recent measured inequality is not a product of high returns on legacy capital, how labor compensation is highest and returns to capital are lowest in liquid, transparent markets where you can go about benevolent business activities without being unduly harassed and without wondering if your property will be confiscated at the next shift in political breezes.
But, the strange thing is that I think Picketty generally agrees with this.
Here is his data, some slides, and some course notes. The subtitle of the course notes is "Is Balanced Growth Possible?", which the notes seem to answer with a qualified "yes". After looking at it for a little while, I thought maybe I had accidentally found course notes of someone using Picketty's work as a starting point for a wider course, but, no the notes appear to be at his site.
It seems like there is a disconnect between the data and analysis he has gathered, and the conclusions and public discussions, including his quote from the book, which I referenced in the previous post. I think others have discussed this, and I'm not qualified to add to the discussion of the econometrics. But, I did have some ideas while looking at these sources that I haven't seen elsewhere, which I will share.
This graph got me thinking about human capital. (Here, Deirdre McCloskey speaks generally about inequality and goes into the topic of human capital briefly during a video interview.<HT: CafeHayek>)
Picketty notes that slavery actually put a price on human capital. In the 18th century, slaves accounted for at least a quarter of measured capital. So, even then, when most workers were either slaves or were earning very low incomes compared to modern standards, total human capital would have dwarfed physical capital in importance. Human capital is basically the dark matter of economics. We know it's there. In the end, it probably is the primary driver of everything that is important. And, we can't see it.
From a financial perspective, human capital is extremely illiquid, so equilibrium returns on it will tend to be high. But, even with a high discount rate, it certainly accounts for a far larger value than physical capital. (In fact, of all the sources, pro and con, affecting inequality today, this is the fundamental driver of increased inequality.) Having a conversation about future inequality 80 years from now, based on returns to physical capital, is like having a conversation about the inflation of the universe without accounting for dark matter.
Oddly, given the public conversation that this has fed, capital, ex-housing, is relatively flat over a very long period of time in the US, and has been falling since the 18th century in Europe, with both roughly settling at around 300% or less of national income. (The fall in relative value of agricultural land is the picture of the disappearing rentier class that I described in part 1. Agricultural land has not become unproductive. It has just been eclipsed. The decline in capital in the UK and France happened because agricultural land was 400-500% of national income in those days of limited access governance.)
The housing portion of these graphs represents the consumption of a small portion of human capital. Using the tools of finance, physical capital can be trade and divided. The risks can be allocated in degrees and in kind to various constituencies. It can even be traded in time, pushing or pulling profits between today and the future.
Human capital is very illiquid. Once you have it, you can't very well sell it. It is time dependent, and it is difficult to trade either within or across time. One obvious trade we frequently make is through student loans, where we take the equity position (ownership of our future "profits") and the bank takes a debt position on the human capital that arises from our education. The subsidies and regulatory concerns around student debt arise because of the illiquid nature of human capital. Debtors don't have a directly associated physical asset to place a lien against.
Probably the main tool that most people use to trade on their human capital is consumption smoothing, and the main outlet for consumption smoothing is home buying. The cultural conventions and financial tools that have developed around housing partly serve to make a trade over time between a young family that needs a larger house when they have low income and low savings but large amounts of future human capital and that same family in the future, when they have high income, high savings, and have expended most of their human capital. The bank serves as an intermediary over time between the present and future family.
We might expect that this sort of activity would represent a fairly stable long term trend. Different generations of families should tend to trade similar proportions of their human capital over time, all else equal. So, the "housing" portion of national capital could be seen as a rough gauge of the growth of human capital. So, in fact, these graphs are demonstrating the continued decline of physical capital as a portion of the total capital held by modern nations.
This does raise some interesting questions. This implies that total capital (including human capital) is increasing in relation to national income (although much more so in Europe than in the US). We might divide homes into equity, which is invested past profits from human capital, and mortgage, which is capital borrowed from the future profits of human capital. Between the end of 1995 and the end of 2005, American households had added $8 trillion in equity to their housing capital and $5.6 trillion in mortgage capital. This was a period when baby boomers were entering the twilight of their working lives. So, this reflects what we might expect. Most of the growth in housing was in equity, which we might expect from an aging population. Compare this to the previous 10 years, when baby boomers were younger, heavy on yet unearned human capital. From the end of 1985 to the end of 1995, mortgages increased by $1.9 trillion while equity increased by only $1.5 trillion.
But, I wonder if some of the 1995-2005 growth in mortgages is related to demographics, too. Even though it is difficult to measure, there must be a complicated web of risk factors in human capital, just as there are in physical capital. One of those factors would be a maturity premium. A young person will have a large amount of potential human capital, but because so much of that capital will be realized decades in the future, the discount rate applied to it would need to be higher. An older worker, however, will have only a few years of human capital remaining to tap, but because of the short maturity, it will have a low discount rate. So, the lifecycle value of human capital will rise in the mid-to late period of a household life cycle because of these sort of capital gains on unrealized human capital. In the market for trading between human capital and real estate, not only would middle aged baby boomers be increasing their equity positions, but their potential future human capital, which they might be trading for mortgage debt, would also be at the peak of its value. This could explain a real estate market that was aggressively expanding in both mortgages and equity, but especially in equity. (Of course, since human capital is so illiquid, there will be a wide variance of outcomes, with some households experiencing extremely low returns on human capital. This is another thing to keep in mind about inequality. Because of the illiquidity of human capital, compared to physical capital, risk sharing is much more difficult with human capital, so it may be, in fact, the declining importance of physical capital that would lead to an increase in experienced inequality. Public redistributionist policies would actually be a more reasonable part of the policy response to this issue than they would be to the supposed r>g problem.)
Here is a graph of a simple model of human capital over a working life. The peak in total present value comes early and is sloped more steeply because of the changing discount value. I submit that this was a factor in the behavior of asset markets in the 2000s, and that we did not manage the descent off of the peak of baby boomer human capital very well....to say the least.
While, in theory, potential dislocations and creative destruction in the marketplace are what lead to potentially inequitable outcomes with regard to unrealized human capital, as the behavior of home equity values in the graph above demonstrates (as would a graph of corporate equities), all capital is susceptible to extreme downside risks. Poor outcomes among physical capital tend to be felt more widely, however, because they are more liquid and more tradable.
The disconnect in the public discussion leading to the concentration on legacy capital as the key to inequality might relate to the measurement issue. Physical assets have a market price, so it is easy to imagine confiscation. And since realized capital gains reflect future earnings, the future can, in effect, be taxed today. There also is not a work-leisure trade-off, so while taxation might lower future investment, it won't necessarily affect immediate production, and thus the negative growth effects are harder to quantify. Human capital can't be easily taxed forward in time, and a tax on human capital results in immediate work-leisure trade-offs.
PS. (added): Hmm. Maybe we are more clever than we realize. Maybe the combination of mortgage subsidies and property taxes ends up subsidizing future realized human capital but taxing realized past income on human capital. This combination of taxes is neutral, so that it wouldn't push households out of the home market. But I think it might serve to encourage human capital development while taxing the actual (and highly variable) gains from human capital progressively. I'm not sure if I can figure out all the implications....
No comments:
Post a Comment