A full PDF version of this blog is available here
In February, the Wisconsin State Assembly passed a bill to provide both child tax credits and a sales tax holiday as a means of refunding a projected budget surplus. The Assembly bill, which modified proposals by Governor Walker, included: a one-time sales and use tax rebate equal to $100 for each qualifying child, and a two-day sales tax holiday in August where the sale of retail products for personal use with sales price no more than $100 would be exempt from the sales tax.
Both parts of the bill were met with criticism. Democratic legislators argued against both proposals, questioning the timing and whether the rebates were the best use of funds from a yet-to-be-realized surplus. Republican leaders in the state Senate have backed the child credit, but have called the tax holiday a “gimmick” and its passage remains unlikely, or at least uncertain. Now, because of differences in the two legislative chambers, the passage of both parts is up in the air.
In this post I discuss the potential impact of both the tax rebate and tax holiday, putting aside the important question of whether the policies are the best use of tax revenue. As relatively small and one-time policy changes, the economic impacts of both policies are likely to be limited. Nonetheless, recent research suggests that a substantial portion, anywhere from 30-66%, of the tax rebate would be consumed. Further, although a primary impact of sales tax holidays is a shift in the timing of spending, recent research has shown that such tax holidays also tend to increase by roughly 30-40% the overall level of spending.
In total the proposed policies should lead to an increase in consumption in Wisconsin of roughly $370 million over the next year, which would lead to an increase in output of roughly the same size. Thus although the policies are not designed for a growth impact, they may lead to a modest increase in output, with a multiplier of slightly greater than 2 from revenue to output.
Consumption and Taxes
The benchmark model for analyzing the impact of taxes on consumption is Milton Friedman’s permanent income theory. This theory states that people base their consumption and savings decisions on their expected lifetime income, and that they use saving and borrowing to smooth out transitory income fluctuations. Since both the tax rebate and tax holiday are relatively small in magnitude and one-time programs, this logic would suggest that tax revenue rebated to households would be almost entirely saved.
While this theory remains an important starting point, over the years there have been a number of well-documented deviations from permanent income theory. In particular, a sizeable proportion of consumers, particularly the young and those with lower incomes, face credit constraints which limit their ability to borrow against future income and smooth consumption. In addition, recent research has emphasized that the same may be true of homeowners who have much of their wealth tied up in the illiquid asset of their home. Thus credit or liquidity constrained consumers would spend a larger share of their temporary tax reduction than the basic theory would suggest.
In addition, the effect of a tax reduction on consumption will differ depending on how it is financed. In particular, a deficit-financed tax cut may signal higher future taxes. In that case, consumers would save a substantial fraction of a cut to pay the higher tax bill down the road. However the tax policies under debate in Wisconsin are driven by expected future surpluses. Thus they reflect, at least in the aggregate, higher future incomes with a fixed tax bill. Expectations of higher income in the future should stimulate higher current consumption, even for consumers who are not liquidity or credit constrained.
This last channel would be altered if the surplus tax revenue was either spent or used for a permanent tax change. If the tax revenue surplus drove an increase in government spending, then consumers would not share as much in the increased income, and thus would face an effectively higher tax burden. On the other hand if the temporary projected surplus led to a permanent tax cut, then taxpayers would expect either these tax cuts to be reversed or spending to be cut in the future. Permanent tax changes would have a larger economic impact than the proposed temporary policies, but they would require future fiscal restraint.
The Tax Rebate
In design, the tax rebate walks a middle path between targeting benefits toward taxpayers in need and providing a broad based refund. The proposed tax rebate is designed as a rebate against sales tax liability, with a flat payment of $100 per child. This lump-sum payment would thus distribute more of the tax revenue to lower incomes compared to an equal-sized reduction in income tax rates, which would give a larger benefit to higher income households. Even more equitable than the child rebate would be a universal per-person rebate. But by targeting the rebate toward households with children, the proposal echoes the recent federal expansion of the child tax credit. Research has shown that the child tax credit and the related earned income tax credit have been powerful tools for improving living conditions for low income households. However while these federal programs (and the existing Wisconsin state EITC) are more narrowly targeted toward low incomes, the proposed rebate applies for all households with children.
The Wisconsin Department of Revenue estimates that 1.22 million children under 18 will be eligible for the credit. According to Census data, there were roughly 608,000 households with children in the state in 2016 (26 % of all households), which means on average two children and thus a $200 rebate per qualifying household. While the rebate would surely be a welcome gain for all families, for many it would not be a sizeable addition to income. In particular, the median income for married couples with children is $95,429, significantly higher than the median household income of $56,811. However for some families the $200 would be much more important. In particular, single female households with children have a median income of $26,803 and a poverty rate of 35.9% (compared to 4.4% for married couples with children). Although not targeted solely to these families, the rebate would provide them assistance.
As mentioned above, recent research has studied the impact of the federal income tax rebates in 2001 and 2008 on consumption. (See also these papers.) Consistent with significant liquidity and credit constraints, these papers found that on average roughly 30% of the income tax rebates were consumed shortly after receipt. In addition, many households used the rebates for down-payments on bigger ticket durable goods like household appliances and cars, and thus the cumulative response was that households consumed 67-90% of the tax rebates.
There are at least two important differences between the federal rebates and the current Wisconsin proposal. First, the federal tax rebates were significantly larger than the proposed $100 per child, which is too small to serve as a down-payment. Thus we would expect the consumption effect to be mostly in nondurables, and not to have as prolonged an impact as the federal rebates. But there is also an important difference in what the rebate signals about future conditions. The federal rebates both happened in recessions, and were at least in part in counter-cyclical policies to try to stimulate consumption and output (particularly the 2008 rebate). By contrast, as mentioned above, the Wisconsin proposal comes in a time of economic expansion in response to forecast tax revenue surpluses. This additional channel of expectations of higher future incomes and tax revenues would lead to a larger consumption response.
Based on these factors, we estimate that households will consume approximately 50% of the tax rebate in total within six months of receipt. Thus the $122 million rebate would lead to an increase in consumption of $61 million. As this increase in demand is too small to significantly affect prices, we would expect it to lead to an equal increase in output.
The Sales Tax Holiday
While the proposed child tax rebate received more broad support, the proposed sales tax holiday faced opposition in the state Senate, which may have doomed the passage of both proposals.
In calling the sales tax holiday a gimmick, Senator Fitzgerald echoed the Tax Foundation, who summarize the arguments against such policies. In particular, they argue that sales tax holidays simply shift the timing of purchases and do not increase overall consumption. In addition, they suggest that sales tax holidays have high compliance costs, discriminate against certain products, are ineffectively targeted to low-income consumers, and distract from more tax effective policies.
However proponents argue that sales tax holidays are an efficient way of providing relief to consumers in periods of high demand. The proposed sales tax holiday in Wisconsin was set for early August and limited to purchases up to $100, in an effort to provide relief form families during the back-to-school shopping season. By lowering after-tax sale prices, this could also help firms by stimulating more sales. The sales tax holiday would particularly advantage consumers who are very price elastic, which tend to be lower income households.
While recent research has found that sales tax holidays do have significant effects on the timing of purchases, they also do lead to an increase in overall consumption spending. For example, a recent study of the Massachusetts sales tax holiday, which exempted items up to $2500, found significant bunching of purchases of large items like furniture, electronics, and appliances on the sales tax holiday. But the overall impact was a 40% increase in consumption on the tax holidays, with no measurable impact either before or after the holiday. Another recent paper analyzed cross-state data for a number of different sales tax holidays, finding a 30-40% increase in consumption from sales tax holidays, with the effects particularly strong for apparel purchases. Further, recent work has also shown that reductions in sales taxes may also stimulate consumption of tax-exempt goods, which suggests the impacts of the holiday may be even broader.
The Wisconsin Department of Revenue estimates that the sales tax holiday would lead to a reduction in tax revenue of $51.5 million on allowable purchases. With the state 5% sales tax, this represents a consumption base of $1.03 billion. As Wisconsin’s sales tax is on the lower end of states, and the purchases are limited to only $100, we anticipate the consumption increase to be on the lower end of the estimated range 30-40%. Thus assuming a 30% increase, the sales tax holiday would lead to a consumption increase of $309 million. The multiplier of 6.0 from tax revenue to consumption is on the low end of the impacts which have been estimated.
While policymakers may question whether the proposed tax rebate and tax holiday are the appropriate use of funds or appropriately targeted, recent research suggests that they would lead to an increase in consumption and output.
Click here to read the full PDF version of this blog.