In the model of the economy as a GDP factory, the most fundamental equation is the production function, Y = f(K,L).
This says that total output (Y) is determined by the total amount of capital (K) and the total amount of labor (L).
Let me stipulate that the economy is legible to the extent that this model can be applied usefully to explain economic developments. I want to point out that the economy, while never as legible as economists might have thought, is rapidly becoming less legible.
For example, the analysis of changes in the trend in productivity requires extremely fine legibility. You start with the measure of Y/L, which is problematic, because you have to add together disparate goods and services to get Y. You have to aggregate all sorts of different specialized workers to get L. Next, to get a trend in productivity, you have to compare Y/L between two periods relatively far apart, say 1986 and 2016. The difference between what you are aggregating then and now is staggering.
Once you look at differences across decades, adjusting for price changes becomes important but impossible. For example, Bret Swanson says that the computing power in his iPhone would have cost $12 million in 1991. If for the purpose of comparing Y/L today to Y/L in 1991 you valued every iPhone at $12 million, you would report an enormous increase in real GDP and hence in productivity.
Finally, to get the change in the productivity trend, you have to attach meaning to the difference of the difference, e.g., the difference between the change in Y/L from 1986 to 2001 and the change in Y/L from 2001 to 2016. That is asking the data to be correct to an additional decimal point.
Here are some other indicators of the decline in legibility:
1. The increasing disconnect between stock prices and earnings. I have made the point about Amazon. In fact, Amazon typifies the decreasing legibility of corporations. Commenters who gave me pushback on Amazon said that they do not think it will end up primarily as a retailer. To me, all the top tech companies have vague business models, particularly if you look ahead several years, and particularly if you compare them with the old industrial giants. Bethlehem made steel. GM made cars. What does FaceGoogle make? Space for ads? Do you think that will still be their primary business five years from now? If so, do you like their prospects? Folks like Ben Thompson seem to think that the business models of all the major tech companies necessarily must evolve radically, which to me makes their future earnings harder to predict.
2. The increasing disconnect between corporate earnings growth and GDP growth. Some of this is due to the fact that a lot of important U.S. corporations are multinationals, so that their earnings are not just a function of what goes on in the U.S. [corrected]
3. The increasing disparities in individual earnings. I bet if you looked at the ratio of the pay of a college president to that of a cafeteria worker today and compared it to that ratio in 1980, it would blow your mind.
4. The increasing disparities in cost of living. I bet if you looked at the ratio of the cost of living in San Francisco to the cost of living in Peoria and compared it to that ratio in 1980, it would blow your mind.
5. Increasing disparities in tastes. The usefulness of a single price index, such as the Consumer Price Index or the GDP deflator, to describe inflation depends on the validity of the assumption of a representative consumer. I don’t see that today. One family shops for groceries at Walmart, and another shops at Whole Foods. One household puts discretionary income into private school for the children, and another puts it into home entertainment.
Still, economists want to treat the economy as if it were legible.
–They tell you that we are in a productivity slump, and we need to explain it and figure out how to solve it. I think that the numbers are not reliable enough to say.
–They tell you that inflation is below target, and central banks are puzzled about how to respond. I think that we see such large changes in relative prices that the very concept of “overall inflation” loses meaning and monetary policy becomes irrelevant until the central banks get into an orgy of money-printing.
–They tell you that “the” real interest rate is low, but what happens if you take health care and education out of the inflation numbers? I mean, if as an entrepreneur you could respond to the scarcity of schools and hospitals (as indicated by rising prices) by borrowing money to launch a new one, then that low real interest rate would mean something. But you can’t do that. Meanwhile, if you’re borrowing to finance a new firm in the solar power business, where prices are going down every year, maybe that interest rate does not look so low.
2. The increasing disconnect between corporate earnings growth and GDP growth. Some of this is due to the fact that a lot of important U.S. corporations are multinationals, so that their earnings are just a function of what goes on in the U.S.
Shouldn’t that be “so that their earnings are NOT just a function of what goes on in the U.S.”?
Yes, I have the soul of a copy-editor.
Arnold, one of your best posts in recent memory. It seems economists often can’t handle phenomena that can’t be aggregated. Statisticians offer many cautions about what happens to data when you aggregate it and use it in a model, but it feels like no one takes those warnings seriously outside the classroom where they are taught.
Agreed – strong post and many valid points. Seems like a much finer-grained analysis is necessary to avoid these problems of aggregation.
Right. However, the trouble is that no one is interested in granular analysis of some tiny sub-field, because there are no big policy implications of any result. Everybody believes themselves to be affected by “the economy” overall, but if you start telling people about the rate of labor productivity growth in the production of alfalfa, then they are going to ask, “What’s that got to do with me?” And policy makers are going to ask, “What’s that got to do with policy?”
The problem is that even if you were able to collect all the little sub-sub-sector grains of economic data (and where you would be able to do more of an apples-to-apples comparison with the past), there probably is no good conceptual way these days to aggregate all those little deltas into big, economy-wide deltas. But that’s what people really want to know about – what’s going on in the big picture.
1. At the household level, there are gaps between “This household is counted as poor in reported statistics” and “How much this household is actually consuming.” I think this is largely because of transfer payments. For some households it probably reflects spending of retirement assets (which are not income).
The consequence is that the statement “This household is poor” doesn’t convey all that much information about the household’s consumption, but rather about the households’ reported earned income.
2. Viewing migration, I propose the hypothesis that historically there used to be a larger percentage of migration (say, between states) that was driven by labor market factors. I think you can argue this just based on possible sources of household income–there used to be fewer opportunities for income / livelihood outside paid work.
Sometime in the last 50 years, a larger percentage of migration has come to result from all kinds of other things–more than 50 years ago. A partial list.
A. Life stage transitions (graduation, college education moves, retirement, becoming infirm and needing family close by).
B. Benefit shopping (Section 8 housing voucher waiting lists are an obvious example).
C. Amenity shopping (hip cities, places of natural beauty; also I would include people trying to avoid dangerous “ghetto” environments at all costs by moving to (say) smaller towns or, paradigmatically, college towns).
D. Ethnic enclaves (Somalis and Hmong come to mind off-hand). I think you could get a handle on this by comparing, for example, immigrant professionals who are dispersed to live almost anywhere in the country by labor markets, and immigrants from certain countries who tend to congregate in a few places.
E. Refugee resettlement policy, which is driven in part by sponsoring non-governmental contracting agencies, and also “chain” migration.
How to prove this second hypothesis I don’t know. It also may be simply that people choose the destination first, and only then do they search for employment, but lack of work opportunities can still reduce the choice set of possible relocations and “veto” certain relocations.
Driver license changes and IRS filings by state would provide data. But you would need time series data–and ideally some qualitative research as well. Need to avoid sample bias, which would be a non-trivial exercise.
Comment on “disparities in tastes” is interesting, especially given the increasing importance of housing in the consumption bundle. The guy in a $300K house in Plano Texas probably prefers his house to living in a $3M Manahattan apartment.
http://easyopinions.blogspot.com/2008/11/political-dictionary.html#macro
The Political Dictionary: Macroeconomics
=== ===
The study of overall statistics about income, debt, production, and employment. You would think that statistics are dull, but macroeconomics is a center of intellectual ferment. Motto: “We just don’t know, but we are willing to guess”. Critics say that it is history confused by mathematics, or mathematics splattered by history.
Macroeconomics is noted for the gigantic cost of the occasional government experiments carried out with fanfare in times of crisis. Experiments in good times are never advertised or acknowledged, for fear of being held responsible for yet again ruining a good thing.
=== ===
The above is merely my view, the discardable view of a peasant. But, it agrees with someone who has credentials in hard science:
=== ===
http://easyopinions.blogspot.com/2009/02/macroeconomics-is-astrology-not-science.html
Macroeconomics is Astrology, Not Science
By Frank J. Tipler, Professor of Mathematical Physics at Tulane University.
[edited] Our leaders are being advised by macroeconomists who haven’t got a clue where they are leading us. Their actions may lead us out of the current recession, or they may lead us into a depression as bad as the Great Depression.
Science is about prediction and precise explanation. It is not enough to construct a different explanation about each past event. Science must produce a consistent, precise explanation for all of the relevant past events.
Then, real science predicts the future and is testable according to those predictions. If a “science” cannot predict what will happen, then it is clear that it does not understand enough about what is going on, and of course it is of no practical use in arranging for a better life.
Real scientists bend over backwards to make their data, methods, and results available for review and criticism. This corrects for personal bias, and allows for quickly sorting out the truth. A true scientist tries to examine all possible explanations for his results, before believing that his new analysis is correct.
=== ===
Of course at any point the economy is illegible? In 2017, even past economies are hard to explain and why did the economy but so hard in 1929 – 1932 or grow so well in 1995 – 2000?
1) In terms of Amazon, it is quite clear some investors trust Amazon to have exceptionally high future earnings compared today earnings. I think it reasonable but I tend to agree with you. In terms of dumb money, I think investor’s trust in Uber is bizarre and trust them too much on their ability to attract drivers while becoming a tech near monopoly taxi company. But of course any reading of economic history, there are all kinds of dumb investments made.
2) Yes, I suspect a lot of this is multi-national corporations but also a well run company with economies of scale is hard to beat.
I’m sorry I don’t have anything of substance to add, but:
This is an excellent whack to the side of the head.
I wish you were in my field.
1. One thing about “the state of the economy” is that it is both an empirical statement about reality but also a suite of potential political narratives full of value judgments regarding fairness and desert and status competition with implications for policy, grabbing for pie-dividing, etc. Different political coalitions are always trying to tell a story in which things aren’t as good as they could and should be for their clients, and the blame goes to the stupid or evil members of the opposition. So, I think there are certainly ways to interpret the essence of the economic phenomena we are observing and to accurately explain what is going on, but part of what is happening is that none of those observations are the useful elements that fit the traditionally effective templates of the kind of stories people want to tell for political purposes. If those old stories seem to be losing some of their political salience it is because they are getting analytically obsolete. Postwar narratives don’t translate well into an Average Is Over world.
2. I think the illegibility concept can be extended to politics and culture at large. A lot of commentators and public intellectuals really seem to be finding the country politically illegible these days, and that’s even when putting aside the Trump surprise. And the reason for this is analogous to the increased granularity necessary to accurately describe contemporary economic phenomenon. There is no longer a good bell curve with some adequately “Representative Agent” being some centrist Median Voter (or Median Household Income Earner). Swing voters aren’t “average” centrists that can be swayed to one side or the other given political developments, but instead people with very volatile political affiliations and which are much more of a distinguishable type that fits a profile unreflective of population averages.
So much seems to depend on multiple variables of identity now, on the trend of the social status of members of ones groups, and the tacit promises of parties or movements to raise or lower that status. It’s hard to hold fissiparous coalitions together when each sub-group is experiencing different conditions and has conflicting interests. The electorate is becomming illegible – when treated as a single thing or just a few major groups – because it is “Coming Apart” in various ways. Indeed, conditions for part of one’s coalition could be trending down while another sub-group could be having it better than ever. It’s hard enough to come up with a single message for these groups, but it’s even harder to really understand what’s happening beneath the surface if one is committed to overarching narratives with a small number of variables and isn’t getting down in the granular data.
I think people in the entertainment or news industries might say something similar. It’s getting harder to give “the people what they want” when “the people” are becomming illegible. And that’s also related to your Narrower, Deeper, Older post.