Second, the effects of tax rate changes are nonlinear in the sense that a small tax increase causes less of a decline in economic activity than a big tax increase. Moreover, the rate of tax that matters most in dampening economic activity is that levied in the last dollar, not the first dollar earned. This is the higher marginal tax rate economists warn about.
These truths combine to explain many things including the Laffer Curve, a hypothesized relationship between marginal tax rates and tax revenues. A staple of supply-side economics this theory roughly suggests that too high a tax rate will dampen economic growth so much that overall tax revenues will decline. The converse is also true that at some point lowering tax rates will boost the economy so much that tax revenues will actually increase. This is because there is both a mathematical relationship (higher rates lead to more revenue) and an economic relationship (higher tax rates leads to less economic activity). While the existence of a Laffer Curve is largely accepted, the best evidence suggests that we are a long, long way from having tax rates that are high enough to feel the effect of the Laffer Curve. So, a tax cut will almost certainly not increase tax revenues, while a tax rate increase will not grow revenues proportionately.
Debates about tax rates are appropriately political. Taxes influence decisions about how much we save, where we live, how many children to have, and how big a home to buy and what type of government we have. We should all understand though, what is true and not true in the debate over the appropriate level of taxes.