Carbon Tax Revenues and Fiscal Multipliers

Resources for the Future issued a report this summer, Deficit Reduction and Carbon Taxes: Budgetary, Economic, and Distributional Impacts. In it, they analyze four revenue-neutral carbon tax proposals: using carbon tax revenues to: (1) reduce capital taxes, including corporate income and personal income for capital investments, (2) reduce personal income taxes and social security taxes, (3) reduce sales taxes, and (4) provide a lump-sum distribution to every adult. They model the effects of each of these separately, and graph the path of the effect of the program on GDP over time, and also, somewhat uniquely, the intergenerational effects of each of the policies, i.e., the effect of each of the four programs on each cohort of people by birth year, reaching out to look at economic effects on the unborn. They then compare each of these with some deficit-reduction proposals, i.e, proposals in which carbon tax revenues simply enter the U.S. Treasury to pay down the national debt.

It is a critically important discussion to have, even if it is politically taboo to even consider a carbon tax that does not somehow recycle revenues. My interest in this blog post, however, is to raise questions about the four revenue-neutral proposals. The conventional wisdom, borne out by this report, is that in terms of effect on GDP, the proposals descend in effectiveness from 1 (capital taxes) down to 4 (annual lump sum distribution to every adult). In terms of trying to reverse the regressive impact of carbon taxes (yes, I am talking about redistribution, that dirty word!), it goes the other way, with 4 going the furthest to not only ease the impact on the poor, but make them better off, and 1 being the most regressive, and in fact making a carbon tax even more regressive.

I try to remain catholic about the value choices implicit in choosing from the 4 revenue-neutral options, and respect the fiscal conservatives that insist that capital taxes are the most growth-maintaining use of carbon tax revenues. But here is the economic question: won't carbon tax revenues recycled back to poor households have a higher multiplier than that recycled back to rich households and corporations? Corporations are currently sitting on huge stockpiles of cash, so if anything, the economic equivalent of disastrously putting cash underneath a mattress would seem to be returning money to corporations. Moreover, we have good reason to suspect that money given to poor individuals gets spent immediately, because of their higher marginal propensity to consume (these people do not have enough as it is) and pumped back into the economy. I am not a macroeconomist, but that sounds right. Mark Zandi at Moody's did an analysis in 2008 looking at multipliers for different uses of federal stimulus money. Below is what his model estimated. I have ranked his hypothetical proposals from highest multiplier to lowest. Proposals highlighted in blue are what I characterize as being close to capital taxes (1), green as labor taxes (2), orange as consumption taxes (3), and red as lump-sum distributions (4):

1.73 Temporarily Increase Food Stamps
1.64 Extend Unemployment Insurance Benefits
1.59 Increase Infrastructure Spending
1.36 Issue General Aid to State Governments
1.29 Temporary Payroll Tax Holiday
1.26 Refundable Lump-Sum Tax Rebate
1.03 Temporary Across the Board Tax Cut
1.02 Nonrefundable Lump-Sum Tax Rebate
0.48 Permanent Extension of Alternative Minimum Tax Patch
0.37 Make Dividend and Capital Gains Tax Cuts Permanent
0.30 Cut Corporate Tax Rate
0.29 Make Bush Income Tax Cuts Permanent
0.27 Temporary Accelerated Depreciation Tax Deduction

The RFF model does not distinguish between households in terms of income or wealth, but rather only has a representative household. This is not to say that using carbon tax revenues to temporarily increase the food stamp program (which has been cut) is literally six times as effective in stimulating spending as a temporary accelerated depreciation tax deduction. But collapsing all households into a representative household seems to me to miss the huge amount of variation in the effectiveness of recycled revenues given to persons of different wealth and different marginal propensities to consume. The pattern shown above seems to bear out what we already suspect about giving money back to people: that money given back to poor people produces a much stronger economic response that money given back to rich people and to corporations.