The quality of scholarship in contemporary academic economics is regrettably low. It focuses too much on complex mathematical theory that makes unreasonable assumptions, and hence proves nothing, while devoting too little effort to describing the empirical reality. Their colleagues in the business schools of the United States typically do a better job of explaining the same things.
A recent, sloppy and inaccurate analysis of effective marginal tax rates under the McCain and Obama tax plans by Greg Mankiw, a Harvard economist, is typical of the discipline's sins.
He concludes, based upon a grossly oversimplified mathematical model, that his true marginal rate is 93% under Obama and 83% under McCain. But, his assumptions about how someone with high income actually invests, and actually makes work or don't work decisions, in the face of the incentives before them, are profoundly flawed. He isn't a tax practioner and it shows. Opportunities to cap tax liabilities abound, and decisions about whether or not to do the work that produces income turn out to be far more "lumpy," and less amenable to incentives, than a conventional marginal income tax rate analysis suggests.