Note: The annual change in state and local taxes excluding property taxes?is regressed on the annual change in the unemployment rate and a time trend. The earlier time-period includes the years from 1959 to 2000, while the later time period includes all years 2000 and after. The lines above and below the estimate represent 95% confidence intervals.

Source:?EPI analysis of Bureau of Economic Analysis National Income and Product Accounts (NIPA) Table 3.3, and BLS Labor Force Statistics from the Current Population Survey.

Copy the code below to embed this chart on your website.

Second, the common claim that property taxes are largely invariant to the business cycle also seems to change once one focuses on more recent data and allows for lagged effects. Property taxes change more slowly as the economy weakens than other taxes, such as sales or income taxes. Property taxes rely in part on assessments of value that are sometimes adjusted only in multi-year increments. Given this, it can take a while for a weaker economy to translate into lower property tax collections. The figure below shows the coefficient for a regression showing the annual change in property tax revenues associated with a 1-percentage-point change in unemployment lagged two years. Before 2000, this coefficient was effectively zero, validating the presumption that property taxes were largely insensitive to the business cycle. But since 2000, this coefficient has become statistically significant and economically meaningful, with property taxes falling as unemployment rises. The post-2000 estimates would, for example, indicate that the average 9.5 percentage points of excess unemployment between the second quarter of 2020 and the first quarter of 2021 would by itself translate into a roughly $60 billion decline in property tax revenues alone within a few years, once lagged effects kick in.

Figure B

Growing cyclical sensitivity of property tax revenues: Annual decline in property tax revenues associated with lagged 1 percentage point increase in unemployment, pre and post-2000

Note:? The annual change in state and local taxes property taxes is regressed on the annual change in the unemployment rate lagged two years and a time trend. The earlier time-period includes the years from 1959 to 2000, while the later time period includes all years 2000 and after. The lines above and below the estimate represent 95% confidence intervals.

Source: ?EPI analysis of Bureau of Economic Analysis National Income and Product Accounts (NIPA) Table 3.3, and BLS Labor Force Statistics from the Current Population Survey.

Copy the code below to embed this chart on your website.

Given these considerations, the finding that state and local governments may face a revenue shortfall of up to $1 trillion by the end of 2021 seems entirely plausible, and federal aid should be on this scale in order to avoid forced cutbacks in state and local government spending from becoming a large drag on economic recovery. Further, unemployment is forecast to be quite elevated even by the end of 2021, and so federal aid should continue as long as economic conditions warrant, and not be set by arbitrary timelines.


* In an earlier post, we found that rising unemployment and reduced economic activity could depress state and local revenues by $500 billion or more by the end of 2021. We now think the number is even higher than this for a few reasons. First, our first estimate was based on a forecast for unemployment (from Goldman Sachs) that was more optimistic than that projected by the more recent CBO estimate. Just the longer and larger rise in unemployment forecast by CBO would push up our estimates. Second, in the earlier post, we used a slightly smaller number for how responsive local government revenue might be to rising unemployment than that used by Bartik. But our findings on the growing sensitivity of property tax revenue to rising unemployment over time highlighted above makes us think his estimate makes more sense. Finally, we based our estimates on the assumption that only state and local revenue shortfalls associated with unemployment rates above 6% would trigger federal aid. But there’s no reason in the current moment to ration federal assistance. The fiscal stress of state and local governments today and for the next two years is entirely the fault of the coronavirus—as a group, these governments were in strong fiscal shape before this shock. Given the huge capacity of the federal government right now to take on debt (as evidenced by historically low interest rates), there is really no reason to not hold state and local governments fully harmless against the fiscal shock of the coronavirus, and this shock can be measured as simply the entire increase in unemployment since the first quarter of this year.

Tagged
黄色网站在线看