We approach a full year since the start of the COVID recession. So, it is timely to assess where the recession has turned out differently that seemed likely at the start. Some of the developments are happy, while others are not.
I begin by noting that employment in Indiana fell throughout 2019. For much of the Midwest, 2019 was effectively a recession year, even if that downturn was masked by the cataclysm of COVID. However, the forces that slowed our economy in 2019 were entirely different from the COVID effects of 2020, and these forces befell a different set of people and businesses.
Most economic forecasts in March and April projected the deep GDP losses in 2nd quarter and rapid recovery in 3rd quarter. I think these forecasts were right enough about the depth and dynamics of the COVID downturn to have been helpful for policymakers and businesses. It appears now that we’ll end 2020 with a downturn that ranks in the top five to seven worst years since the 1920s.
At the beginning of the COVID downturn, most economists argued that the effects of the disease were the causal factors in the recession. With more research we might find that long-term government restrictions slowed recovery, accounting for a small fraction of lost jobs and businesses. But, with the emergence of extensive data on the timing of disease, government actions and business revenue, it is now clear that individual responses to COVID caused the downturn. Government “shutdowns” played no more than a minor role in the depth or duration of the recession. It was always the pandemic.
The economic projections of a deep downturn were accompanied by several estimates of a sustained fiscal catastrophe affecting state and local governments. For many states and local governments, that is turning out to be the case. Local government has lost a whopping 1 million jobs nationally and schools another million. But, two factors have worked together to mitigate the worst likely effects.
The CARES Act actually boosted personal income in 2nd quarter. Most of this boost in spending appears to have been dedicated to taxable sales, and it was all subject to income taxes. At the same time, household spending shifted more heavily to taxable items, like home improvement products, than most economists expected. While the COVID losses are likely the worst post-war shock to state and local governments, the direst potential outcomes may have been averted.
The most remarkable aspect of the COVID recession is the unequal and hidden impacts on labor markets. Too much of the policy discussion has failed to acknowledge how deeply unequal this recession has been. I think the problem lies in overreliance on anecdotal evidence to judge the world around us. This appears to have allowed too many policymakers to ignore deep problems in the official estimates of unemployment in assessing the state of the economy.
The current unemployment rate understates the real level of job losses. A full explanation is too lengthy for this column, but suffice it to say that job losses have displaced maybe three times the number reported in the official unemployment rate. As of mid-November, more than 600,000 Hoosier workers had hours cut so low that they were eligible for public benefits. The low information content of the traditional unemployment data means that the seriousness of the labor market disruptions were too easily ignored. But, I think the real problem is that the workers facing the deepest distress are largely invisible. The truth about their plight is startling.
Nationwide, from January to late October, employment in the top third of workers – those who earn more than $60,000 per year – grew by 1.2%. This is slower than normal, but hardly a recession. Among middle-income workers, those who earn $27,000 to $60,000, employment declined by 4.0%. That is roughly the national experience annually of the Great Recession, in other words, pretty bad. But, among low-wage workers, job losses were at 19%. This is matched only by one year of the Great Depression, and it took four years to reach that share of joblessness.
So, for the most affluent American workers, 2020 was a slow growth year. For middle-income workers, it was the worst year in a lifetime. For the poorest third of American workers, 2020 was a catastrophe akin to the darkest days of the Great Depression. The inequality of this downturn is the most shocking effect of this kind I have ever seen in data. This pattern plays out nearly identically across individual states.
Many of these jobs won’t return. Even if household spending returns to pre-COVID patterns, the acceleration of automation and productivity changes in services has been large and permanent. The only silver lining for workers is that these jobs are geographically distributed across every community. This means we won’t have the concentrated job losses of the 2000s.
The effects of this are already obvious in the data. There are a few sources for student learning in online math classes last year. By the end of the year, when final tests were given, students from families in the top one-third of income made steady progress, with growth of 37%. Middle-income students had almost no measurable change, while low-income students saw their progress down 11%. This worsened over the fall, with low-income kids seeing a 13.7% drop. These effects on students will continue to affect individual and economic performance for a half century.
Doubtless there are many unknown challenges awaiting Congress and state governments as they consider how to deal with the remainder of this downturn. One emerging certainty is that dealing with the unequal effects of labor market and educational experiences need to be a top priority.