Economics Links
Project 2025, Cato version; John Cochrane on consumption and interest rates; James Pethokoukis and Doyne Farmer; Scott Sumner on inflation measurement
Appealing to DOGE, Cato has a wish list. It would get government out of many activities, consolidate government agencies in others, and deregulate in ways that seem consistent with DOGE goals. But it also includes some items that are politically unacceptable and have already been ruled out by President Trump, such as
To better align with longer life expectancies and declining fertility rates, Social Security’s early and full retirement ages should be increased by three years each, to 65 and 70, respectively, and indexed to increases in longevity afterward.
Of course, I agree wholeheartedly with the Cato white paper. But even though I think I am still some sort of adjunct scholar at Cato, I had nothing to do with this effort (or any other Cato effort in recent years). You should at least skim the white paper to see how much opportunity there is to reform government from a libertarian perspective.
If you write a model in which higher real interest rates lower consumption growth, people will look at you really strangely. If you stand up in a FOMC meeting or write an oped and say that higher real interest rates will raise future consumption and output growth, people will look at you just as strangely. It’s interesting that in academic economics we can repeat things without batting an eyelash that nobody in the policy world believes, while the policy world believes a completely different sign, and economists who go back and forth never notice. Economists can say higher interest rates raise consumption growth yadda yadda in an academic seminar, and then explain how higher interest rates will soften the economy going forward (with long and variable lags), exactly the opposite sign, in policy discussions. Cognitive dissonance reigns.
My emphasis. The academic economist’s intuition is that you ask yourself: if I substitute future consumption for present consumption, what kind of deal do I get? The higher the interest rate, the better the deal, so higher interest rates should raise consumption going forward. The policy economist’s intuition is that higher interest rates will lower consumption growth because… (waves hands) …everybody knows that higher interest rates do that.
Meanwhile, I imagine I can hear Scott Sumner ask why an economist would reason from an interest-rate change. The interest rate is an endogenous variable.
Cochrane was asked to write something for a tribute to Bob Hall, and he wound up providing a tour of 40 years of macro. You won’t be able to follow it (I was barely able to follow it), but if you could you would save yourself a lot of trouble wading through dozens of papers.
I’ll give one more excerpt:
Consumption is temporally aggregated. It is the average value for the month, quarter or year in question, while asset returns are point to point. Consumption is seasonally adjusted, which introduces a two-sided seven-year moving average. The theory refers to the flow of consumption, not to the purchase of durables which is really an investment. Consumption services data include housing, healthcare, financial services, and insurance, which most people don't adjust at high frequency. Nondurable goods “last generally less than 3 years” per the BLS, not 1 month. We buy a lot of even ``nondurable'' clothing once a year at Christmas.
The long-term behavior of consumer spending is dominated by the steady increase that has taken place over the years. It correlates with any variable that has a definite time trend. The signal is strong but misleading.
The difference in consumer spending from one quarter to the next gets rid of the misleading correlations with other data. But it is dominated by noise, in part because the distinction between Q1 and Q2 is arbitrary. The signal is too weak to measure any interesting relationships.
Those of you who have read my macro memoir or Specialization and Trade know that I reject everyone’s macro. As of now, my advice to a young economist would be to use AI to create thousands of artificial households and firms, give them different endowments and reasonable heuristics, run simulations, and analyze the results. Speaking of which,
In an interview with James Pethokoukis, Doyne Farmer says,
I think time will tell to what extent this replaces the traditional way of doing economics. I don't think it's going to replace everything that's done in traditional economics. I think it could replace 75 percent of it — but let me put an asterisk by that and say 75 percent of theory.
The traditional way, the MIT way, is to write down a model using mathematical equations. Farmer’s approach uses agent-based modeling. I think if you were asked to choose now, you would pick Farmer’s approach. But academic economics is very path-dependent, and we ended up with the MIT approach as the standard.
I have written about Farmer and also interviewed him.
Economists obsess over whether the CPI or the PCE is closer to the “true rate of inflation”. But how can there be a true rate of inflation if economists cannot even precisely define what they mean by “better”?
…I would have to concede that inflation estimates for a gallon of gasoline or a dozen eggs are far from meaningless. The overall CPI is a hodgepodge composite of meaningless and meaningful data points, all mixed together.
I would add that the goods and services where quality changes matter the most (putting them in Scott’s “meaningless” category as far as inflation measurement is concerned) have over time become a larger and larger share of the economy, so that today they are dominant. So I would strongly agree that measures of inflation are crude. The error bars that one should draw around them are very wide.
If inflation measures are crude, then it follows that measures of changes in productivity are crude. Because you cannot calculate productivity without first making an inflation adjustment in measuring output. I constantly harp on this issue—that the folks making pronouncements about productivity trends are constructing narrative structures on top of quicksand. They are providing substantive interpretations of changes that are within the margin of measurement error.
Scott goes on to say that instead we can measure inflation by looking at wages. But I think that some of the same challenges emerge. If average wages per worker go up by 5 percent per year, by how much has money depreciated? Doesn’t the answer depend on how much productivity has increased, and doesn’t that bring the challenge of measuring price inflation back into play?
substacks referenced above:@
@
With regard to the CATO recommendations, I would add the radio spectrum to the list of federal assets that should be privatized. Public ownership of spectrum is a feature common to Western developed countries, but the US government (the FCC and, for spectrum used by federal government agencies like the Pentagon, NTIA) is arguably the most incompetent spectrum manager in the developed world. It is a fantasy proposal, of course, and public ownership of spectrum is such a sacred cow that even a libertarian think tank like CATO surprisingly, or perhaps not so surprisingly, didn't include it in the list of federal assets to be privatized.
My biggest gripe with academic economists is that the economy spits off an incredible amount of data (in the form of earnings, transaction data, scraped-price data, etc.) that they largely ignore. Investors are much better economists--in that they rigorously observe and model the economy--than economists are.