Fear and Loathing and the Output Gap

The output gap is the difference between what an economy is producing and what it can produce. At the first wave peak of the COVID-19 induced economic crisis, the output gap was initially estimated to be about 15% of GDP at the end of the first quarter of 2020. However, an aggressive Federal “stimulus” program reduced this to a still considerable 10% of GDP by the end of the second quarter of 2020, according to the Congressional Budget Office. Since then, the output gap has continued to shrink, to an estimated 3% of GDP by the end of 2020.

The output gap is a malleable concept that requires two variables, one of which is relatively easy to estimate — how much an economy is producing, and another which is difficult to estimate and always somewhat speculative — how much an economy can produce at full employment.

If economists underestimate the output gap the Federal Government may spend too little and not close this gap and unemployment will be higher than it would be if the gap were closed. If, on the other hand, the Federal Government overshoots the output gap and spends too much, the extent of the overshoot appears not as higher employment but as inflation. If the economy hits its “upper limit” extra demand simply bids up prices for existing production rather than calling forth production that might not otherwise have existed.

In recent years economists have consistently underestimated the output gap. Consequently, many conservative and some liberal economists predicted a surge of inflation after the passage of the Obama stimulus spending measures in 2009. When Trump cut taxes in 2017, many economists once again predicted that the excess spending would just show up as inflation. But in neither case did inflation surge. In the first case, employment fell, but slowly: the output gap was much larger than commonly understood. In the latter case, the stimulus likely spurred growth by about one half of one percent per year and did not result in significant inflation.

Why have economists consistently underestimated the output gap? Why have they consistently underestimated the capacity of the US economy to increase production at full employment?

The entire concept of “full employment” in an age in which American-based businesses can easily tap into overseas labor sources, is itself a somewhat quaint concept. American businesses, even the smallest ones, can often gear up production by outsourcing to an overseas or domestic gig economy, in response to wage pressure in the USA. Part time work, gig and freelance work, often supplements or supplants more traditional labor sources. Consequently the relationships between inputs, chiefly labor, and output, are not as clear as they once were and models based on the history of this relationship have less predictive value than in the past. This has led to a kind of Wild West of output gap predictions — no one really knows.

Consequently, there is little agreement among economists about exactly what will happen when the Biden COVID rescue plan, all $1.9 billion of it, joins with the recently passed $900 million rescue plan, to inject nearly $3 trillion dollars in the US economy. Moreover, it seems likely, that a $1 to $2 billion infrastructure plan will quickly follow the COVID rescue plan, injecting even more demand into the US economy.

This has led to some alarm among many conservative and some liberal economists. One of the best known proponent of the view that the Biden plan is likely to cause inflation is Larry Summers, former advisor to Presidents Clinton and Obama, but, nonetheless, a suspect figure among the left wing of the Democratic Party. Although Summers supports a COVID rescue package of some kind, he thinks that the current package does not do enough to expand the productive potential of the economy.

The more the productive potential expands, the lower the likelihood that Federal rescue spending will result in inflation, since the economy will be able to produce more to meet the newly increased demand for goods and services resulting from the rescue package spending. The Summer argument is not really so much an argument that large sums should not be spent on “rescue packages” or “stimulus” but that the money should be spend more slowly and that more of it should be spent on infrastructure that would expand production.

Indeed, since this infrastructure investment, in his view, would expand production, it would also sustain higher levels of social spending in the future. So, Summers would likely not agree that his argument is not a progressive one but a more prudently progressive one.

The disaster scenario for Summers and like thinkers is that the nearly $3 trillion (and perhaps closer to $4 to $5 billion with an additional infrastructure bill) in Biden-proposed spending will combine with likely strong consumer spending to generate a surge in inflation. Unlike the 2008 recession, consumer balance sheets in the USA are strong. Consumers benefited from prior relief packages and have also cut their spending and enjoyed a run up of asset prices— unlike in some past recessions.

Once COVID restrictions are lifted, they are likely to spend quickly and substantially on restaurants, travel, spectator sports and other long denied pleasures. Most economists, left and right, expect this to happen.

The disaster scenario then would be:: massive stimulation generates so much inflation that the Fed is forced to use its “tools” to bring inflation down. This means increases in interest rates. In normal times higher interest rates are a factor in slowing down the economy and, if high enough, can in themselves induce recession. However, after a long period of very low interest rates, an additional danger is that the withdrawal of very low rates will expose many over-leveraged companies to financial crisis, and that their crisis will enhance any recession.

This result would undercut any further increases in Federal spending so that it would make impossible the potentially multi-trillion dollar infrastructure spending bill that the Biden Administration is purported to be developing. Unlike the COVID relief bill, which is focused on maintaining consumption and combating COVID, the infrastructure bill is a capital goods bill that presumably will expand US production, thereby supporting increased Federal spending with rising productivity and more tax revenue.

However, if the COVID relief bill overshoots and kindles significant, self reinforcing inflation, then all of this will be endangered and, worst of all for Democrats, the recession that all of this portends would arrive at a highly politically inconvenient time, perhaps a few months before the 2022 Congressional elections. In the disaster scenario for Democrats, the Republicans, buoyed by an economy in recession in the latter half of 2022, regain control of one or both chambers of Congress, and then torture Biden with a legislative freeze, until inevitably re-winning the Presidency in 2024 based on their campaign against low growth under Democrats.

Given the serious stakes over the proper measurement of the output gap, it seems possible that the Biden Administration may seek to hedge its bet somewhat, to monitor and interpret possible signs of rising inflation, and to take action to preserve its long term program even if this means delaying or abbreviating in some ways its short term measures. How might this look?

In the short term, the Biden Administration is unlikely to reduce the scope and speed of spending. After all, the output gap may be much larger than many economists expect; indeed, underestimating the output gap has been a ubiquitous practice even for many Keynesian economists for the entire twenty first century. Another factor likely to encourage the Biden Administration to act aggressively is that traditional economic measures — such as the rate of unemployment — have been significantly understated due to the peculiarities of the COVID environment. In February 2020 the labor force participation rate in the USA was 63.3% but this had fallen to 61.4% by January 2021 — a drop of nearly 2%. Much, perhaps all, of this drop is likely to be accounted for by individuals who have dropped out of the work force to take care of children or the elderly and it cannot really be considered a “choice” and therefore not part of the unemployment rate. A more accurate accounting would add this approximately 2% to the current “official” unemployment rate.

In addition, the official rate of business business bankruptcies and closings, or at least the official recording of such bankruptcies or closings, lags their actual physical closing. Many more individuals who are still recorded as having businesses are in fact unemployed. Consequently, the current unemployment rate — 6.7%- is believed by many analysts to actually be over 10%, once COVID-specific conditions are accounted for.

Traditional economic measures are suspect, and past performance at predicting output gaps is suspect even when traditional data is more reliable, but these two factors are not alone in making it difficult to predict the output gap. Uncertainties about the path and development of COVID itself make output gap prediction very difficult. If the new strains of COVID originating in the UK, Brazil and South Africa, take root in the rest of the world, which is rapidly happening, especially for the UK strain, the virus’ impact on US production capacity will increase. Even the threat of the new strains can impact output by necessitating or lengthening lock-downs, or by requiring the development, production and distribution of new or altered vaccines or simply more of or different volumes or configurations of current vaccines.

For example, for most of 2021, the virus, while virulent, has been retreating in the USA, with recorded new case counts falling from approximately 250,000 per day to 90,000 per day. Moreover, by late February 2021, approximately 13% of the US population had received at least one does of a COVID vaccine. The surge of COVID cases that resulted from holiday socializing for Thanksgiving, Christmas and New Years had also started to abate. Consequently, the reproduction number, the statistic that describes how fast the infection is spreading, dipped under one. Any figure under one means that the infected population is likely declining while any number over one means that it is likely growing (the figures are estimates with related error spreads or intervals).

However, the new strains are far more infectious and are rapidly spreading. many infectious disease experts expect the number to rise above one, indicating an increase with the likely result of a new wave of infections in the United States in March and April. This new wave will likely rise well before half the population is inoculated, a goal that is not likely to be met until May.

A new wave would retard business re-openings, lead to further business closings and bankruptcies, further rent arrears, losses of school and work time, and further balloon the output gap. It is perhaps not impossible that virus developments might necessitate further COVID-relief spending and that some forms of this spending could become a much longer term feature of American life than currently imagined.

Yet another reason for optimism about the output gap is that much of the soon to be appropriated $1.9 trillion will be spent on imports (perhaps as much as 15%) and, while in more normal times only about 6% to 8% of the assistance will be saved — consumers, as well as anxious state and local governments — are likely to save a much higher share — perhaps as much as 20%. All of this means that less stimulus will hit the US economy at any one time, but that more of it will have the long term beneficial impact of further strengthening balance sheets and enabling capital investment in imported inputs, and even in enriching trading partners who then buy American goods.

Nonetheless, the enormous uncertainties over the size of the output gap argues for flexibility in the $1.9 trillion COVID relief legislation. It might be too big precipitating inflation and undercutting infrastructure spending. It might be too small requiring further legislation. However, if the currently proposed $1.9 trillion proposal could contain measures which enable policy makers to use the money to meet special contingencies, it could be reserved entirely for COVID relief if needed but shifted to long term infrastructure spending related to long term rebuilding for COVID. For example, it could be used for improved rural internet access or local healthcare delivery modalities.

The reality is that we are all guessing about the output gap, making educated guesses but unable to garner enough reliable data to firmly predict how bad the endgame for COVID will be, or even if we are near the endgame. We also don’t really know how much of the rescue or “stimulus” money will be saved or spent; nor do we really know the real impact of the COVID experience on American productivity. Our companies and even our non-profits have just made an enormous investment in online work management that may very well more than compensate for losses of businesses during the pandemic and lead to big productivity increases.

But we are wandering in the dark. The situation calls for determination (a BIG rescue plan and an infrastructure bill) but with the flexibility to spend the funds over longer periods of time should this prove necessary.