Debate in the Senate about extending the extra $600 unemployment insurance (UI) benefit beyond this week’s expiration left 32 million people hanging while senators took off for the weekend and millions of unemployed contemplating how they might make their August rent or mortgage payment.

Conservative Republican senators and the White House have seized upon a narrative that the relief program they approved in March has some people making more money on unemployment than they earned on their jobs.  On the face of it, the concept goes against our historic beliefs about safety net benefits – they should not come too easy for the recipient since that would discourage effort and motivation to be productive.  But these are not normal times. Following beliefs about human behavior that made sense in different times could be disastrous for the COVID-19 economy, currently in shambles.  See Barron’s article from Michael C. Klein, “32 Million Reasons Congress Must Extend Jobless Benefits”, (paywall likely).   We add a few more reasons to the discussion.

  • To get aid to the suddenly unemployed from the full-stop economy in March, Congress approved what looked like the quickest, most simple approach.  The program was initially designed to help the unemployed achieve 100% of the average worker’s weekly salary, around $1000.  This was a “blunt tool” to get money out to the millions of laid off workers as simply and quickly as possible.  However, using an arithmetic mean resulted in those who made less than the average worker getting more money while those that made more than the average worker getting less money. Peter Ganong, Pascal Noel and Joseph Vavra published a paper on this topic, that was picked up by the New York Times, FiveThirtyEight and other outlets.  They also offered a calculator to show UI replacement rates at different earnings levels. 
  • Turns out, there are many laid off workers making less than the mathematical average.  As Ganong, Noel and Vavra point out, median pre-COVID-19 wages are considerably lower than the average.  Low wage workers have suffered disproportionately high layoffs in this unusual economic downturn.  (This is also one reason that average wages have seemingly gone up in recent labor reports.  They haven’t actually gone up; its just that COVID-19 effects washed out so many low wage workers from the calculation.)
  • The U.S. job market has gone through significant changes in its composition since the 2008-2009 recession.  Many higher paid manufacturing jobs left the US employment base, to be replaced by lower paid service workers – particularly in people-facing leisure and hospitality jobs and people-facing healthcare jobs.  Interestingly, the Ganong, et al paper shows the highest replacement rates among food service workers.  They also show that states with the highest replacement rates are among our poorest states: Mississippi, Alabama, Georgia, Arkansas, Kentucky, West Virginia, North Carolina, and a few others.  Changing to a proportional replacement rate would likely hurt people in these states.   
    • In June, 2019, there were 17.2 million people employed in the “leisure and hospitality” (LH) category or 11.3% of the employment base.  This category includes the arts, entertainment and recreation, performing arts and spectator sports but importantly too, food service, drinking establishments and accommodation such as hotels, motels and the like.  More than 12.3 million were employed in food service and drinking establishments in June, 2019. 
    • By  April, 2020 the LH category had fallen by half, to 8.5 million.  The food service and drinking establishment categories alone had shed 6.2 million employees.
    • Reopening measures in May in many states led to exuberant rehires, and in particular, food service and drinking establishments added 3.2 million employees between April and June. 
    • The story might have a happy ending there, but rapid acceleration in COVID-19 cases led nine states to reverse their re-openings.  In most cases bars have been re-closed and restaurants were mandated to severely reduce indoor dining and overall capacity. The New York Times is keeping state-by-state track of re-closings and pauses here.  Charts at the Johns Hopkins University COVID-19 trackers also show state-by-state cases relative to re-openings.  We will likely see these changes reflected in higher unemployment when the July figures come out (just before Congress takes its summer vacation). 
  • We detail these points since these workers rely heavily on tips for compensation.  In January, 2018, the National Employment Law Project (NELP) came out with a report asserting that wait staff rely on tips for 56.5% of their compensation while the proportion for bartenders is 54.%.
    • While we do not have exact figure on how much of this non-wage compensation was counted in the study of “replacement” unemployment insurance, we suspect that W-2 wages likely understate total compensation and overstates the replacement rate.  The NELP report was produced in reaction to lawsuits that employers were taking away tips that restaurant and bar workers were supposed to receive.  In addition, some states allow employers a “tip credit” to permit below minimum wage payments.  These are just a few reasons employers and employees may understate tips. This is a complicated issue with tax implications as well, no room here, but worthy of further study.
    • The growing courier and messenger category of workers also includes jobs that are compensated with tips, as do the Uber, Lyft and taxi drivers that have become a popular form of transportation in the gig economy.
    • Shorting people in this category of employment the extra $600 UI benefit may not rock the economy given their low-income status but could create dislocation and suffering among millions. 
  • Another argument adopted by the Senate Republicans and the White House is that being paid more than prior earnings may in some cases make it difficult for employers to hire or hire back unemployed. This is largely faulty thinking.  In the Barron’s article Michael C. Klein stated that “there isn’t evidence of it in the data…it’s possible that the extra income makes it harder for businesses to hire workers by discouraging some from accepting low-paying offers. But these are trivial complaints compared to the program’s effectiveness at sustaining household incomes and spending during the pandemic.”  Klein cites research from former US Treasury economist Ernie Tedeschi.
  • Further, the decision to return to work is not based solely on the monetary benefits.  Given the idiosyncratic decision-making around school openings, unemployed parents have to make tough choices about returning to work apart from just the monetary calculation.  Also, people in households caring for medically compromised individuals may have lack confidence that their employers offer sufficient protections to keep them safe from spreading the virus to family members. Economists at the Becker Friedman Institute also took up the childcare point in an April paper. 
  • On the plus side, the extra income improved consumption in May and June, which as Barron’s article points out, feeds back positively through the economy.  Also see another paper from the Becker Friedman Institute about the surge in consumption, particularly among lower income households. 
  • One more thing.  It is unfair to those essential workers who are also in the low-wage categories.  Many remained employed and packed your groceries, your Amazon purchases, made the deliveries and at the front line, healthcare workers took care of the sick and dying.  Certainly, they too, should get a bonus to bring their compensation closer to the U.S. average.  Think about what that might do for consumer spending!   

As the Senate debates reinstating expired UI, they are looking at setting a percent of prior income to “fix” the replacement problem.  This would give more money to higher earners and less money to low earners.  It would also take time to implement and slow down checks.  State unemployment systems are already overwhelmed with the sheer volume of applications.  Their systems are antiquated.  Making the change would entail reprogramming, of course, and introduce a higher probability of error in an already strained system.  You can find feature articles in nearly every newspaper around the U.S. about people that have not yet received the first round of unemployment checks.  To understand the complexity, take a look at Matt Darling’s twitter thread for a state-by-state rundown of complex state UI formulae.

We hope congressional actions will not send us into reverse, but we’re not optimistic.  Frequent articles express the concern that there will be a wave of homelessness and dislocation as the eviction moratoria on rent and mortgage payments expire,  here, here, here, and here.  It is possible that the alarm about evictions is overstated for now, since some people may have deferred their payments to conserve resources and because they could – they will likely pay-up rather than face eviction.  But what happens in a few months when the extra savings run out, they are still unemployed and receiving lower benefits?