The trend towards incorporating data into policy decision-making is a noble one and has helped many governments and companies develop smart growth policies.  However, misreading the data either inadvertently or when it supports a pre-determined conclusion can lead policy-makers away from facilitating economic recovery.  We see at least three examples, discussed below.

“Windfall”:  A recent Wall Street Journal editorial (paywall likely) described a “windfall” that state and local governments receive caught our attention. The so-called windfall relies on some faulty assumptions.

The WSJ cites recently released Census data on state and local revenues to support their premise.  We looked at the same data and drew different conclusions.  First and foremost, most states revised their individual income tax filing and payment deadlines to July 15th, 2020 to be consistent with the federal income tax postponement.

That said, comparing Q3 2020 (July, August, September) with Q2 2020 or Q3 2019 would, indeed, show much higher collections, but these are misleading comparisons.

A better comparison would be between Q3 2020 and Q2 2019 to capture the quarter in which taxes were filed in both years – notably, in 2019, Q2, the April, May, June quarter.

Doing this shows, in fact, a decline of about $10 billion in the aggregate state individual income tax collections across the 42 states that have some form of individual income tax – from 2019 to 2020.  Income taxes made up about 43% of aggregate state tax collections in Q3 2020 Census figures and 42% in Q2 2019.

To be sure, this is a rather noisy set of data since many states made changes to their state tax rates and tax brackets in 2019 (reflecting the strength of 2018 markets and economy) and again for January 1, 2020. Some states lowered their rates or changed the relative tax burdens across brackets in the context of a nearly full employment, a strong and growing economy. 

Comparing the two large income tax collection quarters shows a wide range of results when considering individual states. Worst off with triple digit revenue reductions are Connecticut, New Jersey, Utah and Wisconsin.   At the other end of the spectrum, seven states had double digit increases in individual income taxes between April 2019 and July 2020, including Arkansas,  Vermont, Louisiana, South Carolina, Georgia, Virginia and Colorado (more reddish than bluish). 

The Tax Foundation has a thorough recap of tax changes by state, found here.   

State and Local: Comparing aggregate figures for both state and local income tax collections introduces the Census Bureau’s seasonal adjustment. The Census Bureau indicates that state figures are not seasonally adjusted (as in the figures we cite above), so presumably the differences between adjusted and unadjusted figures relate wholly to local collections.

The seasonal adjustment is mysterious to us, and we invite our wonkier readers to provide insight. For example, adjusted figures show a $40 billion increase in income tax collections between Q2 2019 and Q3 2020. Unadjusted aggregate state and local figures show a drop of $12.1 billion for the same quarters.

Looking at 12-month figures, ending September 2020 and 2019 also show adjusted and unadjusted figures moving in different directions. Adjusted state and local income tax collections went from $451 billion in 2019 to $465.8 billion in 2020, an increase of about 3% as the WSJ reports. On a non-adjusted basis, state and local income tax collections fell from $454.8 billion in 2019 to $450 billion in 2020. We are skeptical of the accuracy of adjusting for “seasonality” during the pandemic and prefer the non-adjusted figures as a better economic indicator. As far as we know, seasonally adjusted dollars are not yet an acceptable form of currency for consumption or wages and we certainly hope that the pandemic does not become a routine seasonal factor.

To wit, the new stimulus package will undoubtedly help move the economy forward through early spring but our federal leaders ignored the role that states and local governments play.   Money for individuals will trickle up to state and local coffers in some ways and aid to school districts will help make schools safer to re-open and expand modern technology for students in low income households. But how can business be effective without clean water, reliable electricity, well-maintained roads and transportation systems, and an educated workforce?  Add to this, modernizing election technology and processes and unemployment insurance systems. These are services paid for and provided by state and local governments, not the federal government. 

Wage Gains:  Second, average wage gains tend to get press when data are released, but these data, too, often miss some basic math. Ten months on, there are still 10 million unemployed in the US. While unemployment is down from highs in April, it is still twice as high as pre-pandemic. A high proportion of those unemployed were in lower-paid hospitality professions.  Eliminate this population from the average and, voila, the average goes up.  Notably, an increase in average wages based on a smaller, higher paid population does not increase state/local revenues but rather reduces overall withholding taxes. The on-going issues of the unemployed burden state unemployment systems, county welfare systems, and economic activity in general.  The conclusion that “increased” average wages mean that  state and local government finances are fine misses the bigger picture.   

Unemployment Replacement:  There was a study showing that unemployment benefits in the first CARES Act replaced more than 100% of pre-pandemic employment for those at the lowest income groups.  These findings prompted a repeated congressional narrative (notably from Sen. McConnell) that providing too much money serves as a disincentive for the unemployed to return to work.  But this narrative misses a key factor in the back-to-work decision for many. During normal times, the incentive argument carries water, but the pandemic also resulted in school closings.  In some states, schools were re-opened and then closed again as the pandemic swept through communities that were seemingly untouched in the first wave. The BLS reported in its newsletter and the Kansas City Federal Reserve Bank in its “Main Street Views“, that pandemic unemployment disproportionately affected women, noting that there are about 15 million single mothers in the US.  With schools closed, and limited, affordable and safe childcare available, the choice to return to employment (to the extent jobs are available) is about more than just dollars.   

Is it time to re-visit federal incentives?  For one, the structure of unemployment insurance assumes there are enough jobs available if one only looks hard enough and long enough.  But when other institutions that support return-to-work break down, “looking hard enough” may no longer be the most productive feature.  Perhaps a targeted childcare program for unemployed parents would make more difference than short-term, blanket “helicopter” money.  Perhaps revival of “Teach America” or a “Vista” program could employ out-of-work Generation X,Y,Z-ers and furloughed teachers.  One could mirror the “pods” that more affluent, work-from-home parents set up to facilitate home-schooling and care for their kids. This, paired with good broadband, WIFI and computers for those that cannot afford them could help both unemployed parents and their kids maintain and improve skills during this time.  For sure, too, holding state and local governments harmless for pandemic induced losses would also prevent job loss, create new jobs and support businesses with modern and efficient infrastructure.