Ukraine Payroll Tax Cut: Can It Work? 

The Ukrainian parliament has approved a bold law dramatically cutting a payroll tax: from 41.6% to 16.4%. The main motivation behind the move is obviously brutal recession and an attempt to lure business in paying taxes, rather than ship their money abroad. Furthermore, the Ukrainian government hopes that the revenue lost due to reduced tax rates will be compensated by an increase in the tax base. Is such a plan reasonable?

One can use simple calculations to assess the plan. Let R be the reported taxable income, t is the tax rate so that the tax revenue is t*R. The reported taxable income is equal to the true earned income (W) minus evaded income (E): R=W  ̶  E. Denote the share of evaded income E in total earned income W with M=E/W.

The elasticity of tax revenue with respect to tax rate t is given by ɛ_TR=(dR/dt)*(t/R). After simple algebra, one can find a simple formula for this elasticity:

ɛ_TR = 1 + {  ɛ_W*(1/(1  ̶  M))  ̶  ɛ_E*(M/(1  ̶  M)) }

where ɛ_W=(dW/dt)*(t/W) is the elasticity of the true earned income with respect to the tax rate, ɛ_E=(dE/dt)*(t/E) is the elasticity of the evaded income with respect to the tax rate. These elasticities measure how sensitive the behavior of the private sector is to tax rates. ɛ_W is often interpreted as a measure of the supply response. One can expect ɛ_W to be negative (if tax rates increase, people work less) and ɛ_E to be positive (if tax rates increase, tax evasion increases too). These elasticities have the right signs: if the expression in curly brackets is smaller than -1, then a reduction in the tax rate will generate more tax revenue! So the key question is how big the combined effect of ɛ_W and ɛ_E is.

I am not aware of estimates of ɛ_W and ɛ_E for Ukraine but there are estimates of these elasticities for Russia when the Russian government implemented the flat income tax in 2001. The Russian estimate of ɛ_W is close to -0.05, that is, a ten percent reduction in the tax rate increases earned income by only 0.5 percent because people will work harder. This is a fairly weak response but many estimates of the supply response suggest low sensitivity. The estimate of ɛ_E for Russia is about 0.4. In other words, a ten percent reduction in the tax rate increases earned income by four percent. The response of evasion to tax changes is much larger than the response of actual labor effort in Russia. While there is no consensus estimate for the share of evaded income, the range of estimates for Ukraine is 0.3 – 0.5. Thus, M is likely between 0.33 and 0.5. When we put these ingredients together, we obtain

ɛ_TR = 1 + {  -0.05*(1/0.66)  ̶  0.4*(0.33/0.66) } = 0.73

The positive magnitude of ɛ_TR suggests that this tax cut is not going to be self-financed. It will reduce the tax revenue by ɛ_TR time the percent change in the tax rate: ɛ_TR* (41.6%  ̶  16.4%)/41.6% = 43%.

This result is based on the Russian estimates. What would it take to match the hopes of Ukraine’s government?

First, the size of the shadow economy may be larger. As M (the share of evaded income in earned income) gets larger, ɛ_TR gets closer to zero and hence the tax cut could be revenue neutral. If M=0.5, which is probably an upper bound for Ukraine, ɛ_TR=0.5 and, consequently, the revenue from the payroll tax should decrease “only” by a third.

Second, people may work harder (create new business, work longer hours, increase labor force participation, etc.) than implied by the estimates for Russia. Saez et al. (2012) provide a critical review of the estimates of elasticity of taxable income for developed countries (this is a different object than ɛ_W but since in developed countries tax evasion is much less widespread, one can take it as a proxy for ɛ_W): it ranges between -0.12 and -0.4. If we take ɛ_W = -0.3, the implied ɛ_TR=0.35 and the revenue falls by only 21%. Note, however, that the short-run response of the supply side to the tax change may be weak because it takes time to start a business, find a new job, move to a new location, etc.

Third, the private sector in Ukraine may show much stronger tax compliance than its Russian counterpart. For example, the government could give not only a carrot (reduced tax rates) but also a stick (prosecute evasion more aggressively; this did not really happen in Russia). While it is hard to guess a plausible magnitude in such a case, suppose the government can double the elasticity of the tax evasion response ɛ_E to 0.8. In this case, ɛ_TR=0.53 and the revenue falls by 32%.

Now consider a favorable scenario: the shadow economy is large (so that the capacity to expand the tax base is large), people and businesses really work harder (so that the size of the pie available for taxation increases), and the private sector strongly complies with the law (so that a bigger share of the pie is reported to tax authorities). Specifically, if M=0.5, ɛ_W = -0.3, ɛ_E=0.8, we obtain ɛ_TR=-0.25! The negative sign of this elasticity means that a tax cut will generate more revenue. In other words, the previous tax rate (41.6%) put the government on the wrong side of the Laffer curve.

Obviously, these calculations are oversimplified and omit many important details but they highlight several key points for the government if it wants to make this reform to work and generate revenue.

First, the government has to allow businesses and workers operate as freely as possible to maximize the “supply” response and make ɛ_W as big as possible. The law promises to simplify tax administration which helps in this dimension.

Second, the government should couple the reduced tax rate with increased pressure to comply. This may conflict with the first point but a few high-profile cases where tax evasion is punished may be enough to enforce tax discipline. Fortunately, the law stipulates increased fines for evading taxes by a factor of 1.3 to 3.

Third, the short-run response may be anemic because of adjustment costs and so the government has to be patient. Some clauses in the law may help accelerate the adjustment. Specifically, the tax cut is temporary: the low rate of 16.4% applies only for 2015; after that the rate is set at 25%. Typically, business try to take advantage of temporary tax cut by shifting their activity to periods with low tax rate. So this is probably be a good thing in the short run. However, the law could also dampen the response of “slow” margins such as getting new educations, finding a new job or occupation, etc. because the adjustment costs for these margins could outweigh benefits of a one-time gain from this deep tax cut in 2015.

Fourth, to minimize lost revenue from the tax cut, the tax cuts should apply only to marginal income. In other words, if firm A pays all taxes, a tax cut for this firm does not change the behavior of the firm (it continues to pay taxes) and shifts resources from the government to firm A. There is no gain for the government.  In contrast, if firm B switches to the legal side of the economy, the government gains the revenue. So the tax cut should be structured in a way such that it applies only to firm B. There are provisions in the law aimed to achieve this. For example, to get the low rate, a firm must increase its wagebill by 30% (one can think of this as requiring the combined elasticity of ɛ_W and ɛ_E to be at least 30%). However, this magnitude is not enough to compensate for lost tax revenue (mechanically, the increase of the wagebill has to be 60% to keep the revenue the same given a 60% reduction in the tax rate). In addition, with inflation close to 20%, the cost of increasing wage bill by 30% for firms is not that high so that even infra-marginal firms like firm A could qualify for the tax cut.

Fifth, a tax cut for one source of income and keeping (or increasing) rates for other sources of income can lead to income shifting when workers and businesses simply re-label their costs and income to take advantage of lower rate for one of the income sources. What was capital income before, could become labor income now. As a result, the revenue from the payroll tax may increase at the expense of other taxes.

In the end, a self-financed tax cut is a pretty rare thing. Indeed, the calculations suggest that it will take an unusually favorable coincidence to have the tax cut revenue neutral. Fortunately, the law has some provisions that should help to make it work (increase penalty, affect marginal firms, temporary additional reduction of the tax rate in 2015, etc.) Therefore, given the large project fiscal deficit, the government should prepare a plan for how it is going to cover the shortage of the revenue. This radical tax cut is risky.