The tax cut expectations game

I was reading the Penn Wharton Budget Model’s dynamic analysis (this is an outfit run by a George W. Bush-era Treasury tax guy) of the tax bill Monday night, and I saw something interesting in its modeling assumptions. It had to decide what to do with the fact that the bill, as written, has a bunch of tax cuts expire in order to conform to the Byrd Rule, even though GOP leaders say they don’t actually want them to expire.

The way Penn Wharton handled that was to “model the dynamic (economic) effects of the amended bill as households and investors expecting no sunsets prior to the sunset dates,” but then later “to be consistent with the actual bill, the sunsets then unexpectedly expire as scheduled.”

This is by far the most favorable way to assess the bill, because assuming the tax cuts are permanent gives you the full growth-boosting impact, but then letting them unexpectedly expire means you don’t actually have the revenue loss. Despite this very generous scoring methodology, Penn Wharton says that on a dynamic basis, “debt increases between $1.9 trillion and $2.2 trillion” — i.e., by more than the static $1.5 trillion loss — because this is basically not a very good tax plan.

But I also think the whole exercise helps explain some of the gap between theoretical models, in which low taxes supercharge growth, and practical real-world experience, in which this rarely seems to pan out.

As the Penn Wharton explanation makes clear, the supply-side growth-boosting impact of tax cuts comes from expectations about the future. Lower taxes make new investments more profitable, which leads to more investment, which leads to more growth. The difficulty here is that while it’s easy to write down an equation in which economic actors come to expect permanently lower taxes, it’s difficult to write a tax bill that actually accomplishes this.

Consider that in 2003, Republicans passed a tax bill whose centerpiece was a cut in rich people’s capital gains and dividend taxes. But in 2010, Democrats folded into the Affordable Care Act a special 3.8 percent Medicare surtax on rich people’s non-payroll income — including dividends and capital gains taxes. And in the winter of 2012-'13 those 2003 tax cuts expired, so 10 years after passing the tax cut, it turned out that taxes on rich people’s investment income were higher than they were at the end of Bill Clinton’s term in office.

The tax bill Congress just passed, meanwhile, though it contains plenty of good news for rich people it does not specifically reverse the impact of the 2010 and 2013 hikes in capital gains and dividend taxes. So on net, if you reacted to the second round of Bush tax cuts by lowering your estimate of what the long-term dividend tax rate would be, you would have been making a mistake.

Which is just to say that the future is hard to predict. My best guess of what the corporate tax rate will be in 2027 is influenced by the passage of a corporate tax cut this week, but honestly it’s not influenced by it all that much. If you’re making an investment decision for which foreknowledge of what will happen in the 2018 and 2020 elections is make or break, well, then you are trouble. Maybe taxes will be way higher in the future because Trump will spark a huge backlash. Who knows?

These days, nobody likes preachy blue-ribbon commissions, and bipartisanship is a total nonstarter in Congress. But if you actually want to generate supply-side tax effects, you need to create an expectation of permanence. And bipartisanship seems like one of the only plausible ways to accomplish that.

This is an abbreviated web version of The Weeds newsletter, a limited-run policy newsletter from Vox’s Matt Yglesias. Sign up to get the full Weeds newsletter in your inbox, plus more charts, tweets, and email-only content.