ANSWER. I haven’t read the Stiglitz book, so I will have to answer from general principles. Let me know if Stiglitz somehow has a way around the problems I will point out.
In simple economic models taxing capital has one of the biggest long-run negative effects on the economy of any tax. It looks OK in the short run, but with lower investment, the capital stock gradually declines. In this spirit you would be better off taxing land a la Henry George, since the amount of land won’t decline even if you tax it. But taxing the buildings on top of the land is like taxing any other kind of capital. (However right now we tax houses very lightly compared to factories, so if it weren’t for the housing bubble’s aftermath, we would be better off taxing houses more and factories—which employ people—less.)
One way to tax capital some in a way that won’t hurt capital formation is to shift from labor taxation (such as Social Security taxes) to consumption taxation, since in the long run the shift to consumption taxation increases taxes on people who have the wealth to consume more than they earn. But in the shorter run, the shift to consumption taxation hits people who are temporarily having to consume more than they earn. Also, consumption taxation has a big life-cycle element. The biggest category of people who have the wealth to consume more than they earn is senior citizens.
To return to the negative effects of capital taxation, the short-run temptation is much like the short-run temptation to have rent control. In the short run, the supply of apartments is inelastic, so rent control looks like a pure transfer from landlords to tenants, but in the long run rent control is a disaster because it makes it unattractive to build new apartments or even keep the old apartments in good repair. The stories of evil landlords not taking care of apartments that pop up under rent control are oh so predictable from economic theory.
At the beginning I said that in simple economic models, taxing capital has one of the most negative effects of any tax. That is true if the tax on capital is constant. If the government taxes capital now and promises never to tax it again, the story gets more interesting. In theory, forcing all companies to issue non-voting stock to the government worth 90% of a firm’s value would have no distorting effects, and so would be the perfect tax as long as people believed the government would never do it again. But if the government will do this once, what is to stop it from using the same logic to do it again. This is called the “dynamic inconsistency” problem.
Getting government institutions set up to block the recurrence of this ultimately self-contradictory logic behind taxing capital a lot now and promising never to do it again is actually one of the trickiest problems implicit in “Leveling Up: Making the Transition from Poor Country to Rich Country.” I think often here of the history of England. Confiscating accumulated wealth was always attractive to the king. Only the evolution of limitations on the King’s power to take accumulated wealth (the essence of the “one-time” capital tax) in the end allowed England to get the development of factories that helped it to become rich.