Home arrow Blog arrow Problems with the Laffer Curve

Problems with the Laffer Curve

Bold claims are often made for the Laffer Curve and the related Laffer effect. They are usually made by people who favour lower tax rates.

Laffer Curve

The Laffer Curve is often represented something like the image above (courtesy of Wikipedia) and the theory associated with it is briefly described by Wikipedia as follows:

In economics, the Laffer curve is a theoretical representation of the relationship between government revenue raised by taxation and all possible rates of taxation. It is used to illustrate the concept of Taxable Income Elasticity (that taxable income will change in response to changes in the rate of taxation). The curve is constructed by thought experiment. First, the amount of tax revenue raised at the extreme tax rates of 0% and 100% is considered. It is clear that a 0% tax rate raises no revenue, but the Laffer curve hypothesis is that a 100% tax rate will also generate no revenue because at such a rate there is no longer any incentive for a rational taxpayer to earn any income, thus the revenue raised will be 100% of nothing. If both a 0% rate and 100% rate of taxation generate no revenue, it follows that there must exist a rate in between where tax revenue would be a maximum. The Laffer curve is typically represented as a stylized graph which starts at 0% tax, zero revenue, rises to a maximum rate of revenue raised at an intermediate rate of taxation and then falls again to zero revenue at a 100% tax rate.

It's worth emphasizing that this curve is derived as pure theory with no empirical justification. Subsequently, claims have been made that actual data supports the theory, but results are ambiguous. There are theoretical objections.

Everyone can agree that the graph must go through the origin. Any tax levied at a rate of zero will have a yield of zero. That is the end of the agreement.

Beyond the origin, we do not even know that it is right to suppose a continuous curve. It is not impossible that the graph could have discontinuities (that is to say, it may be impossible to draw it without taking the pencil off the paper).

Nor do we really know what happens at a 100% tax rate. The theory glibly assumes that people would not work and therefore the yield would be zero. But it makes no sense to extrapolate our present system, with nothing changed except the rate of tax. Obviously, it would be absurd for people to complete tax returns or for employers to keep tax records.

Assuming a tax rate of 100% is meaningless without a much fuller description of the circumstances that are being envisaged. It certainly doesn't follow that people would not work without a financial inducement. In fact, there is plenty of evidence to the contrary. Many activities that could use employees are actually carried out by volunteers. A huge amount of human activity is voluntary, and much of it is indistinguishable from similar activity carried out by paid employees.

But the fundamental issue is that if people did not have incomes from working, then we would have a radically different society, and it would make no sense to describe it in terms of a curve relating tax rates and tax yields.

The situation is further complicated if we look at a wider range of taxes. The original idea was framed in relation to income taxes. But its enthusiasts have tended to assume that it applies to any tax whatever. Suppose we take capital gains tax, for which a Laffer effect is frequently claimed, although never proved. It is not at all obvious that a 100% rate of tax would give a zero yield. It is perfectly possible that people would organize their affairs with the aim of receiving income and avoiding capital gains. But it might well be the case that despite their planning, capital gains did sometimes arise, and with a 100% tax rate, the yield would not necessarily be zero.

So now we have a situation where the graph of tax yield against tax rate goes through the origin, but we do not know any other points on it, nor do we know that it is continuous. A further point, not yet mentioned, is that in addition to not knowing that the graph is continuous, we have no idea what its shape might be. Unless you assume what needs to be proved, there is no reason for assuming that the graph will be a curve similar to the illustration shown above.

There may be more than one local maximum, so even if empirical evidence were found that showed the tax yield tending to fall at some particular rate, it does not follow that further increasing the rate could not find another, higher, local maximum. There is nothing in theory (nor any evidence) to support any particular shape of graph.

The Laffer effect refers to the fall in total tax yield that will arise if there are downwards sloping portions of the graph. However, we have now established that we do not know enough about the graph to know whether there are any downwards sloping portions, so we cannot deduce purely from theory whether there is a Laffer effect.

When it comes to empirical verification, supporters of the Laffer curve generally offer anecdotal evidence that is vulnerable to numerous distortions. What happens in an actual economy is affected by very many factors. Even what happens to a particular tax is affected by issues as simple as advanced warning and expectations about future changes. For example, the US is often cited in relation to capital gains tax, but since notice is normally given of changes and frequent changes have led to an expectation of higher rates being later reduced, the realization of gains tends to be moved to minimize tax take. That has no bearing on what would be the tax take if a higher rate were applied permanently. Many proponents of the Laffer curve for capital gains tax also fail even to take account of the rudimentary issue of varying availability of capital gains according to the state of financial markets. In general, attempts to provide empirical support for the existence of a Laffer effect are so poorly carried out as to lack conviction.

Finally, proponents of the Laffer Curve and Effect frequently go on to make highly speculative claims about how different taxes effect the economy as a whole. Again, these lack an adequate evidential basis. There is, in any case, room for argument about what should be the objective for economic actions. For example, Hyman Minsky advocates prioritizing full employment over economic growth, and it is not irrational for a society to be unimpressed by an economic arrangement that benefits only a tiny minority.

#128008 •  2 November 2012 8:49pm by Martin Brampton • Vote: Agree (309) Disagree (217)

blog comments powered by Disqus