Index the Payroll Tax to Unemployment?

I’ve been spouting for a long time that the Earned Income Tax Credit (our largest means-tested cash transfer program) should be indexed to unemployment, creating an automatic stabilizer in the economy.

But Traders Crucible has what I’ve decided is a far better idea: index payroll taxes to unemployment — decreasing them during downturns and bringing them back up when the economy’s strong. He sort of sidles into the idea via some fairly wonkish posts here and here, with help from frequent MMT commenter Beowulf (one of the main people behind the trillion-dollar platinum coin idea). His latest is here — pointing out that the idea is getting play from none other than Peter Orzag, former director of the Obama Office of Management and Budget and currently vice chairman of global banking at Citigroup.

Here an attempt to explain the logic in plain language.

First, to state what many will consider to be obvious — feel free to skim these five points:

1. Every reasonable policy expert (those who don’t think economies are perfectly self-correcting at all times) agrees that government should engage in countercyclical policy (monetary, fiscal) — loosening money, spending more, and taxing less in downturns where unemployment is high and there’s lots of unused (undemanded) capacity, the reverse when the economy is strong and demand for products, services, and labor is banging against supply capacity, potentially driving inflation. I would venture to suggest that most everyday folks understand and agree with this slightly counterintuitive idea, at least in the abstract.

2. Political dynamics tend to encourage exactly the opposite, as we’re seeing right now: a weak economy and the resulting weak revenues/high deficits frighten short-sighted politicians (and less-clever voters) into austerity at the worst possible time. Likewise when things start to boom: politicians feel flush because of high revenues so they don’t feel pressure to raise taxes or cut spending (“take away the punchbowl”). This is espeically a problem on the fiscal side, but as we saw so clearly in the Greenspan era, the Fed can be equally feckless with monetary policy during upturns.

3. One important cure for that dynamic is automatic stabilizers: programs that automatically increase government spending and/or cut taxes in downturns, the reverse in upturns. The main ones, currently, are unemployment insurance, Medicaid, food stamps, the Earned Income Tax Credit, etc.

4. The Fed obviously engages in very intentional and decidedly discretionary countercyclical monetary policy. Its relative political independence (emphasize: relative) allows for that, but its dual mandates (inflation, employment) and its intimate connections to and concerns for the financial industry tend to hamstring it when it comes to spurring employment.

5. On the fiscal side, the political world often succeeds in implementing some discretionary, non-automatic countercyclicals (recently including the 2% cut in payroll taxes, extension of unemployment benefits, and the Making Work Pay tax credit, all of which are expiring). But 1. they almost always fail to rein things in on the upside (Clinton and a Republican congress partially excepted), and 2. the political process, especially in the current landscape, makes those initiatives fitful and rather arbitrary (and judging by the current state of things, insufficient) — hardly the inspirers of confidence and continuity that people and businesses like to feel before they spend their money on consumption and investment.

Which brings us back to the payroll tax idea. Why is it a good idea to index it to unemployment? What’s wrong with the idea? Why is it a better or worse idea than indexing EITC to unemployment?

Here in no particular order:

Once the policy is in place, politicians don’t need to expend any political capital for the stimulus to occur. It just happens when unemployment rises. And it automatically goes away when the economy is good.

Unlike all (?) our other automatic stabilizers, it works via tax cuts, not spending. This not only makes it more politically viable, it diversifies and expands our portfolio of automatic stabilization policies, potentially giving a higher chance of their collectively spurring growth in downturns.

Adjustments to the payroll tax are already on the political table — they’ve actually been used this very year — so this is not just a pie-in-the-sky idea that will never happen.

Tax cuts are of course always attractive to Republicans, making it more likely that this idea could pass the Norquist test. (The problem, of course: what goes down must come up. Will they ever agree to the second part?)

The policy directly encourages employment (people to work, businesses to hire) via transparent and immediately apparent incentives — actual cash money, every week, right on the paycheck or payroll — not via mysterious, poorly understood, (extremely) controversial, multilevel monetary or fiscal transmission mechanisms.

In its simplicity, it’s easy for politicians and voters to understand. (As long as they can grasp countercyclical economic policy…)

The machinery is all in place, on the public and private sides. Adjust the (already very simple) payroll deduction percentages, and it happens. (I recently set up a six-employee business on Intuit Online Payroll; it took some doing to change over, but it now costs $19/month to run payroll with direct deposit and most tax calcs/deposits happening automatically — ten minutes’ work every two weeks. And, key: Intuit always has the current deduction tables. The businessperson doesn’t have to think about it. The same is true for other payroll systems.)

There’s no need to set up a government jobs program (as recommended by many progressives and MMT proponents) with the huge overhead that would inevitably be involved; lower payroll taxes in downturns would drive employment in the private sector instead. (This is not a panacea: 1. It’s not clear whether the incentives/stimulus would be sufficient to lift us out of recession. 2. there’s damned strong evidence that we need a lot more provision of public goods — infrastructure, education, basic R&D, etc. Private employers won’t provide those goods, no matter how much employment is subsidized. But that’s arguably a different discussion — how much should government be doing? — from the automatic stabilizer issue.)

It’s not clear who should get the tax cut — employers or employees or both. But whichever, and in whatever proportions, the realities of tax incidence will provide incentives on both sides (at least over time) via wage negotiations. Given the stickiness of wages, though, it’s unclear how quickly frequent changes in the deduction percentages could ripple out into wage changes, hence transferred incidence. This, along with political expediency, leads me to suggest that employees and employers should probably get equal cuts.

It’s not clear what the formula should be for linking payroll taxes to unemployment, or how frequently the deduction tables should be adjusted. I leave that to greater minds than mine.

Defenders of the Social Security trust fund will point out, rightly, that payroll tax cuts starve the trust fund, weakening its financial position and hence its power as a political tool to protect Social Security, and potentially revealing it as an instance of political and accounting legerdemain. But as I’ve pointed out, the trust fund is something of a progressive own-goal in any case, and should perhaps be put out of our misery.

Since the payroll tax is only paid on earned income below $110,000, the bulk of the cuts would go into the hands of lower earners, who are more likely to spend rather than save, hence spurring demand from producers/employers, who will ramp up supply in a virtuous cycle.

Since the payroll tax is regressive — falling largely on those making less than $110K per year, and only on money that people actually earn below that level — cutting the tax would result in a more progressive tax system overall. This is appropriate and fair given the massive and increasing disparity in income and taxes (and especially wealth) over the last thirty years, and more progressive taxation also seems to be more economically efficient over the long term — resulting in faster economic growth. (Or at least, no slower growth.) Unfortunately, that progressivity would go away in the good times when payroll taxes are raised back up.

Payroll tax cuts would not be as progressive as EITC increases, because payroll tax cuts go to all “earners,” not just low-income earners.

An increase in the EITC would actively incentivize employment, as opposed to removing a disincentive. Money is fungible so they arguably come to the same thing (depending on how the arithmetic works out), but active subsidy is both more targeted (to lower earners) and potentially more “persuasive.”

EITC is already too complicated (it should be simplified [PDFs]), resulting in a large number of errors on tax returns and some cheating. Adding regular changes to the rates or eligibility requirements would potentially make it completely unworkable.

Payroll tax cuts, since they appear on weekly paychecks, would have higher “salience” than EITC changes — which only appear to taxpayers annually — resulting in stronger and more immediate incentives to work and hire. (There was an Advance EITC program that allowed employees to get the credit on their regular paychecks, but 1. For whatever reasons — lack of publicity? making an already complicated system even more so? — less than 1% of EITC recipients signed up for it, and 2. Congress ended it last year, presumably as a dodge to avoid shifting costs forward into the then-current fiscal year.)

There’s lots more to say on this, but I think those are the main issues.

Thoughts?