The ‘Representative Agent' Is Always Rational. The Rest of Us Are Not.

The economists call this person a "representative agent," who is as real as the Easter Bunny or Abominable snowman. The representative agent represents us all in a model that shows how the economy functions: 335 millions Americans are reduced to just one. As originally conceived the representative agent was fully rational, unemotional, and forward-looking, with perfect knowledge of all relevant facts. Like nobody you know.

Since the introduction of representative agents in economic models more than fifty years ago, many economists, including some of the most prominent names of the profession, have criticised their lack of realistic assumptions.

In 1992, Alan Kirman of the European University Institute, Florence, Italy wrote in The Journal of Economic Perspectives that it was clear that a "representative" agent deserved a decent funeral.

Paul Romer of New York University indicted among others the representative agent in 2016, in a scathing criticism that began: 'For over three decades, macroeconomics have gone backward.

In 2018, Joseph Stiglitz of Columbia University wrote in The Oxford Review of Economic Policy that 'the central problems of finance - bankruptcy, debt, and asymmetrical information - simply cannot be solved by a model of representative agents'.

Over the years, there have been several attempts to make the model of the representative agent more realistic. One such attempt was the introduction of the financial sector. One fix was to add some consumers that aren't very forward-looking, and live paycheck to paycheck. The basic idea is still in use, partly because the math and understanding are easy, and partly because nothing else was available.

Now, something new is emerging. Heterogeneous Agent New Keynesian, or HANK is what it's called. There is no representative agent in a HANK-model of the United States. Each agent has their own unique characteristics, such as rich or poor, lender or borrower, etc. This heterogeneity allows us to easily see the distributional impact of an increase in government expenditure, for example.

The New Keynesianism (NK) part of the acronym is the main theory of modern economics. It assumes people are essentially rational and that markets are largely efficient, but it leaves room for imperfections. For example,'sticky prices' that do not change instantly in response to changes in demand or supply. HANK has a high chance of being widely accepted by the profession because it is closely linked to New Keynesianism.

Last week, I had a conversation and an email exchange with two of the researchers in this young field of HANK. They are Benjamin Moll, professor at London School of Economics and Political Science and Matthew Rognlie assistant professor at Northwestern. It was an interesting look at a way of thinking that may become mainstream in the coming decade.

Rognlie wrote to me: 'I am actually surprised that this hasn't received more media attention.' 'I guess that in academia, we are always over-excited by the latest developments. But I think this one could be quite significant.'

A brief history of macroeconomics is first. John Maynard Keynes demonstrated in the 1930s that government deficit spending could help an economy come out of a recession by putting money into people's pocket. Robert Barro, building off the work of Robert Lucas, who died in May, argued in the 1970s that this trick would not work, because the taxpayers would realize the deficits had to be paid, and they would then save for future tax bills, thus canceling out the stimulus.

To deal with this so-called Lucas criticism, economists tried to build macroeconomic models based on micro-foundations. These are how individuals behave within the microeconomic realm of spending, savings, working, and leisure. To make the models more tractable, economists averaged all individuals into one farsighted person, our famous representative agent. The dice was cast.

The heterogeneous agent model has gradually penetrated. Initially, they were used by economists to examine the distributional impacts of a tax reduction or interest rate increase. In the 1990s, they were used to estimate macroeconomic factors. Early practitioners include Per Krusell of Stockholm University and S. Rao Aiyagari of University of Rochester, as well as Ayse Imorohoroglu at University of Southern California Marshall School of Business.

Moll says that the interest in what is now known as HANK grew following the global financial crises of 2007-9. This was something conventional economic models could not have predicted. He said that it was only in the last five years when papers began to appear. The acronym was created by Moll along with Greg Kaplan, Giovanni Violante and two other co-authors in an article published in The American Economic Review in 2018 entitled 'Monetary policy according to HANK'

Moll stated that the newer theories are more in line with economic reality than older ones. The newer theories are also more closely related to behavioral economics which acknowledges that people make poor decisions because they're impatient and rely on rules of thumb.

The representative agent, if the government issues $1,000 checks, will look at the infinite horizon. He or she will prudently decide to deposit almost all the money into the bank. Heterogeneous agents, on the other hand, behave differently. The rich are able to bank government money while the people living on the fringes spend it all. This second-round impact is that the spending impulse leads to companies hiring more people, increasing output and so forth.

Bottom line, fiscal policy generates growth much more effectively in a HANK than in a model with representative agents. Rognlie stated that HANK models accurately predicted the impact of the pandemic stimuli, which contributed to the high inflation we see today. He said that a HANK-model also accurately predicted high energy prices to weaken Germany's economic growth by suppressing non-energy consumer spending.

Rognlie, Adrien Auclert, and Ludwig Straub have published a paper using a HANK-model to demonstrate that excess savings in an economy 'trickle' up to the accounts of wealthy people.

HANK gives better results in terms of monetary policy. In a representative agent model of debt, it's not really a problem: People borrow and others lend but the end result is a wash. HANK, however, shows that interest rate increases benefit lenders, while harming borrowers. Since borrowers live more on the edge, they reduce their spending more than lenders increase it.

Moll explained that HANK was made possible by two changes. The first is a greater computing power. Another is academic training. To really understand models and program them, you'll probably need a Ph.D. within the past 20 years.

Moll stated that HANK was'still a baby, but it is catching on'. Moll said that the Federal Reserve Bank of New York, Norges Bank (the central bank of Norway), and other banks, such as the European Central Bank are experimenting with HANK. He said that he didn't believe it would completely replace existing models. It will be a model they can use in addition to the existing ones.

Michael Englund, Benjamin Engen and their outlook

In a Friday client note, Michael Englund, Benjamin Engen, and Action Economics, Boulder, Colo. wrote that the latest Bureau of Labor Statistics employment report -- which showed that the economy had added 209,000 new jobs in June, and that the unemployment rate was down to 3.6 percent, -- demonstrates that "overall, the economic is continuing to show resistance to the ongoing tightening of monetary and fiscal policies since 2021." The Federal Open Market Committee of the Federal Reserve 'will have no choice' but to increase the federal funds rate during its next meeting, scheduled for July 25-26. They also added that they'may' raise it again at the September 19-20 meeting. The Federal Reserve's Federal Open Market Committee 'will have little choice' but to raise the federal funds rate at its next meeting on July 25-26 and 'perhaps' at the September 19-20 meeting as well.

Quote of the day

The genius of high-dynamism was the restless spirit that conceived, experimented and explored throughout the economy, leading to insight and luck. This grass-roots attitude was fueled by the new values and attitudes that defined the modern age. To return to high dynamic, it will be necessary for these modern values to prevail over the traditional ones.