The Big, Permanent Tax Increase Inside the Tax Cut Act

Let's talk about tax-bracket indexation.

Let's take a closer look, shall we?

Photographer: Zach Gibson/Bloomberg

It was not a signature element of Ronald Reagan's 1980 campaign 1 or of the tax-cutting plan his administration unveiled in 1981. Several Republican lawmakers (among them Senate Finance Committee Chairman Bob Dole of Kansas) pushed hard for it, and Reagan agreed to include it in the Economic Recovery Tax Act of 1981, albeit with a delay until 1985 to mute its negative budget impact.

Over time, though, this indexing of tax brackets to inflation has become arguably the most significant and lasting consequence of the 1981 tax legislation. It's certainly been the most expensive, with a revenue impact that as far as I know is no longer tracked but by this point must be multiples larger than any other element of the 1981 cuts. It also seems to have been the inevitable and right thing to do. As the New York Times put it in an editorial in 1984, when some lawmakers -- and Democratic presidential nominee Walter Mondale -- were pushing to delay or alter the provision's looming onset:

Indexation is the fairest tax reform in many years. To tamper with it is to betray a trust.

So anyway, you'll never guess what the legislation formerly titled the Tax Cuts and Jobs Act of 2017 2 does! It tampers with indexation! I wouldn't exactly say that the change is betrayal of a trust, but it is a tax increase that will get bigger and bigger over the decades, and it will weigh heaviest on those in lower tax brackets. Unlike most other individual income tax changes in the new law, it is permanent, and given its positive revenue implications, it is hard to imagine any future Congress rolling it back.

This sure seems like it's worth a closer look. To start, let's go back for moment to the original debate over tax-bracket indexation. Some lawmakers in both parties objected, with reasoning that today seems more than a tad weird. This is from a 1981 Times article:

Senator John H. Chafee, Republican of Rhode Island, said that indexing would make taxpayers indifferent to the need for slowing inflation and that consequently it would be harder for the Government to muster the political will to slow it.

"We are going to have a whole new class of citizens who shrug their shoulders at inflation and say, "It doesn't hurt me,'" Senator Chafee said.

Senator Russel [sic] B. Long, Democrat of Louisiana, opposed indexing on the ground that it would "make inflation worse by pumping more money into circulation at a time inflation is at its worst."

Yes, John Chafee was the father of later Rhode Island senator and governor (and 2016 Democratic presidential hopeful) Lincoln Chafee. Russell Long, who had been chairman of the Senate Finance Committee for 15 years until Republicans took over the Senate majority in 1981, was the son of (in)famous Louisiana populist governor and senator Huey Long. 3 I'm going to give Long a pass because he was far from alone in failing to foresee that the inflation rate would plummet from 12.4 percent in 1980 to 4.3 percent in 1985. But I think Chafee's argument may well have been diametrically wrong.

That is, indexing of tax brackets and of government benefits such as Social Security (which was linked to the consumer price index starting in 1975) probably did make inflation less of a concern for voters. But it also made inflation much less attractive to lawmakers and other government officials. Without indexing, inflation brings a steady decline in real government benefits and a rise in real tax rates. In other words, it makes it possible for Congress to pass tax cuts and add new government programs without busting the budget. This worked brilliantly in the 1960s, when inflation averaged 2.6 percent a year; in the 1970s (average inflation rate: 7.2 percent), Social Security recipients and taxpayers -- aka voters -- began to notice what was going on. The gig was up, and after inflation indexing was implemented for benefits, tax brackets and, in 1997, some Treasury securities, the political rewards to inflation declined. Perhaps not entirely coincidentally, inflation did, too.

There were concerns, though, that the consumer price index maintained by the Bureau of Labor Statistics wasn't measuring inflation correctly. This seems to have first come up in the measurement of gross domestic product. Inflation-adjusted or "real" GDP is the most watched measure of economic growth, and adjusting GDP with an inflation index such as the CPI that's based on a fixed basket of goods probably skews it. I'll let Federal Reserve Bank of St. Louis economist Kevin Kliesen explain:

First, the structure of the economy -- meaning the relative prices and types of goods and services produced -- changes significantly over time. For example, think of the advent of the Internet and the products and services now offered online. Second, these relative price changes cause corresponding changes in the purchasing patterns of consumers. If, for instance, technological innovations lower the cost of producing a product, which should then lower its selling price, the quantity demanded of that product should increase and, accordingly, its importance in the calculation of GDP should increase.

In 1992, after years of discussion and research, the Commerce Department's Bureau of Economic Analysis 4 began calculating an alternative GDP with a "chained" measure of inflation that attempted to reflect these changes in economic structure and consumer purchasing patterns. In 1996 it made that the main measure of real GDP. This move was controversial in itself, with skeptics insisting ever since that U.S. GDP growth is now overstated, but it was the concurrent report by a panel of economists led by Michael Boskin, a Stanford professor and former chairman of President George H.W. Bush's Council of Economic Advisers, the final version of which was submitted to the Senate Finance Committee in December 1996, that really stirred things up.

The Boskin report, titled "Toward a More Accurate Measure of the Cost of Living," estimated that the CPI overstated inflation by 1.1 percent a year. It also recommended that the BLS "move away from the assumption that consumers do not respond at all to price changes in close substitutes" and concluded that "Congress and the President must decide whether they wish to continue the widespread overindexing of various federal spending programs and features of the tax code." In response, in August 2002, the BLS began publishing an alternative inflation measure called the Chained Consumer Price Index For All Urban Consumers, or C-CPI-U, and ever since then politicians in Washington have been talking about switching to it for the indexing of spending programs and tax brackets.

QuickTake What You Need to Know About ‘Chained CPI’

The closest they came to taking such a step before this year seems to have been in 2013, when, as part of a budget overture to congressional Republicans, President Barack Obama proposed indexing Social Security benefits to chained CPI. That didn't go over super-well with Obama's fellow Democrats, and he dropped the idea in 2014. Now, however, congressional Republicans have inserted chained CPI into the tax code -- and don't be surprised if they try to force it on Social Security next year. 5

Is this a travesty? No, not really. The argument that chained CPI is a better measure of the cost of living than the unchained version seems pretty legit. Still, switching to chained CPI does have consequences. In the case of the tax code, it amounts to a stealthily growing tax hike -- increasing taxes by just $800 million next year but by $31.5 billion in 2027, according to the Joint Committee on Taxation. The total increase over the next decade is $133.5 billion, in a bill that cuts taxes by $1.5 trillion over that period. In the decade after that, switching to chained CPI will bring in almost $500 billion in additional revenue, estimates the Urban-Brookings Tax Policy Center. And the amount will just keep going up.

As someone who has complained repeatedly about the negative revenue implications of the tax bill, I guess I shouldn't gripe too much about one of its most fiscally responsible provisions. But it does seem important to note that, because of the way tax brackets work, the tax increases brought on by chained indexing will be bigger in percentage terms for those near the bottom of the income distribution than for those at the top. According to a 2013 analysis by the Tax Policy Center, taxpayers in the second income quintile (currently those with cash incomes of $25,000 to $48,600) would see the biggest percentage increases, with their tax rate going up by 0.4 percentage points and their after-tax income declining by 0.4 percent 15 years after a switch to chained CPI. Meanwhile, those in the top 0.1 percent of the income distribution (incomes above $3.4 million) would see average percentage changes of effectively zero.

It's not a huge tax increase for anybody, unless you extrapolate into the next century. But it is a tax increase, it is regressive, and it is probably never going away.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
  1. It did rate a brief mention in the 1980 Republican Party platform.

  2. After a parliamentary-rules challenge by Senate Democrats on Tuesday, Republicans had to go with the formal title of "An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018."

  3. Subject of possibly the greatest political biography ever written, a famous Robert Penn Warren novel (as a thinly disguised "Willie Stark," who was played in the movie version by Broderick Crawford), and also this Randy Newman song.

  4. The BLS is part of the Department of Labor.

  5. Fun fact: Social Security benefits are indexed not to the all-urban-consumers' CPI-U, the headline inflation measure and the one used until now in the tax code, but to CPI-W, the Consumer Price Index for Urban Wage Earners and Clerical Workers. That's because in 1972, when Congress voted to start adjusting Social Security benefits for inflation, the CPI-U didn't exist yet. Interestingly, if Social Security were indexed to CPI-U instead of CPI-W, benefits would be slightly higher than they are now.

To contact the author of this story:
Justin Fox at

To contact the editor responsible for this story:
Brooke Sample at

Before it's here, it's on the Bloomberg Terminal.