I want to throw out an idea of how to encourage "labor hoarding" by employers -- maintaining payrolls even with decreased demand -- in the face of what I fear will be a serious recession. Others have begun to talk about this, including @arindube , @SHamiltonian & @econjared
I preface this by saying that this proposal is not intended to be a substitute for OTHER proposals to encourage more sick leave, help state/local governtment budgets, or shoring up UI, as has been proposed by my colleagues Chris O'Leary & Steve Wandner. upjohn.org/research-highl…
Nor is it necessarily a substitute for proposals to encourage more work-sharing, as my colleague Sue Houseman & Katherine Abraham have proposed in past, & which makes sense now. hamiltonproject.org/papers/encoura…
Nor is this a substitute for proposals to boost overall demand via cash payments, as argued by @jasonfurman & @Claudia_Sahm
The proposal is to revise a form of the Job Creation Tax Credit proposal that John Bishop and I made back in 2009. That proposal was for a 15% tax credit for employers that boosted payroll over their payroll in the same quarter of the preceding year. epi.org/publication/bp…
The revised proposal would be a 15% credit against employer payroll taxes for payroll that exceeded 90% of the payroll in the corresponding quarter the prior fiscal year. The proposal would go into effect in Q2 2020, and apply thru Q42020, & look back to Q2 2019 for Q2 2020, etc
Note that this would apply to any employer paying payroll taxes, so it includes non-profits and state and local governments
Now, in our proposal, we assumed a certain elasticity of response by lowering marginal labor costs with this credit, based on prior research. We ended up with a cost per job created that in 2020 dollars is $35,680 per job created.
Key point: this is much cheaper per job created than most fiscal stimulus, which costs generally over $100K per job created. See my article from the @UpjohnInstitute newsletter in 2010: research.upjohn.org/cgi/viewconten…
Now how much would this proposal cost for the rest of 2020? The Penn Wharton budget model has scored President Trump's proposal of completely eliminating payroll taxes, with a combined rate of 15.3% on both employer & employee side, as costing $807 billion.
The proposal I am outlining is far cheaper because it only gives a 15% credit for a small portion of the payroll. It's trying to lower the costs of maintaining payroll on the margin, to encourage labor hoarding, which is intended to raise impact and lower costs.
Now it might seem that the cost should be 15%/15.3% times 10%, since I am only applying the cost to the last 10% of payroll. But that is not quite right since I am applying the credit to the excess of over 90% of one year ago, and some firms will be expanding in recession..
Based on the BLS's Business Employment Dynamics data, in recent years the gross job gains due to expansions and openings are a little over 10% of base employment: bls.gov/bdm/
If we ignored that some firms go out of business or contract so that they are ineligible for any credit, then the effective percentage of credit would be around 15% times around 20%/110%, to reflect these gross job gains. But this is an overstatement.
This could be much better scored by CBO or someone with access to microdata on firms. For moment, I assume that on avg we are applying this 15% credit to around 15% of payroll. Then the cost, as % of cost of Trump proposal would be (15%/15.3%) times 15% or a little less than 15%.
So cost of proposal would be a little less than 15% of $807 billion cost of Trump proposal, or around $119 billion. If the cost per job created is around $36K (see above), the proposal would create around 3.3 million jobs, compared to what we otherwise would have had.
If we wanted to cut cost of proposal, we could restrict it to employers with less than 500 employees. According to Census Bureau, around 41% of private payroll is in such firms, so this would cut cost in half.
Would this cut effects in half? Maybe not, if one assumes that small business are more responsive than large.
I should also add that my paper with Bishop argued that if job responsiveness was what we outlined, the effects of the proposal on tax revenue and spending needs would offset about 85% of gross fiscal cost, so net fiscal cost far less. Not a Laffer curve, but lower net cost.
Now, one could rerun this proposal under varying assumptions about: responsiveness of employers to this incentive; size of floor for credit; credit %, etc.
2 more things to add. First, none of this is to say that a full payroll tax cut is cost-effective. In my view, the $807 billion payroll tax cut would create less jobs than what I am proposing, because it wouldn't lower employer costs as much.
(Some economists would say the employee side payroll tax would be shifted, but I don't think this would happen in SR).
Second, I think a proposal such as this might have "norm" effects, as argued in latest book by @econnaturalist , that might exceed what could be captured by ordinary economic models. We are trying to set norm for employers: let's all maintain payrolls & labor hoard.
Finally: this is an idea for discussion. I am sure there are ways to improve this proposal. For me the question is: is there something we can do now to encourage employers to maintain or even expand payroll, that we can do at affordable cost per job this would save or create?
