Politics & Policy

The Chained CPI: A Bad Deal All Around

Tax hikes and benefit cuts should be made rationally and openly, not with a stealthy statistical switch.

The Chain-Weighted Consumer Price Index (or chained CPI, for short), which President Obama included as part of his formal budget proposal, seems like a no-brainer for any White House–GOP grand bargain on the budget deficit. After all, the chained CPI is a better measure of inflation than the indices the federal government currently uses, and this simple technical fix would reduce entitlement spending and increase tax revenues by a combined $340 billion over ten years, providing something for both sides to like and dislike. Yet as pressing as federal deficits and debt are, the chained CPI is bad policy that both liberals and conservatives may come to regret.

The Chained CPI is a better measure of inflation than the current indices because, instead of simply measuring the changing price of a fixed basket of goods, it recognizes that purchasing decisions change when prices change. For instance, if the price of apples falls, people will buy more apples and fewer pears, and vice versa. This holds true regardless of which fruit is more expensive to begin with.

Controlling for these changes yields inflation rates around 0.3 percentage points lower than those calculated with the fixed-basket approach. The Bureau of Labor Statistics states that the chained CPI “is designed to be a closer approximation to a cost-of-living index than other CPI measures.” And the BLS is right — but it’s not as simple as merely getting a more accurate measure of inflation. With both Social Security and the tax code, a better inflation measure could lead to worse public policy.

In Social Security, the chained CPI would replace the CPI-W (intended for urban wage-earners and clerical workers) in calculating annual cost-of-living adjustments (COLAs). Once fully implemented, lower COLAs would reduce a retiree’s average lifetime benefits by around 4 percent, cutting Social Security’s long-term shortfall by around one quarter.

Yet while Social Security does need to be fixed, and lower benefits for middle and high earners should be a part of the equation, smaller COLAs weaken a feature of Social Security that actually works: The program’s generous inflation adjustment counteracts the absence of inflation adjustment in private pensions. And unlike most reforms, which reduce benefits progressively — meaning that the poorest pay the least and the wealthiest the most — COLA reductions fall hardest on the oldest beneficiaries, who are most at risk of poverty. An 85-year-old is 66 percent more likely to be in poverty than a 65-year-old, but the chained CPI will cut the 65-year-old’s by only 1 percent and the 85-year-old’s benefits by 8 percent (since the 85-year-old will have 20 more years of having his payment shaved). Moreover, the chained CPI, like CPI-W, doesn’t account for the fact that older retirees spend disproportionately on health care, a sector in which inflation is particularly high.

A better policy would peg COLAs to wage growth, which is around 1 percentage point faster than inflation, coupled with a lower initial retirement-benefit level to keep lifetime receipts the same. The lower starting benefit would dissuade workers from retiring too early. Higher benefits later in life would focus resources where the danger of poverty is greatest, as well as compensating for the fact that most non–Social Security sources of retirement income aren’t inflation-indexed at all.

And despite reducing payouts to retirees, the chained CPI wouldn’t come close to making Social Security financially sustainable. While policymakers have focused on health care, Social Security’s 75-year shortfall rose from $4.3 trillion in 2008 to $8.6 trillion in 2012. Adopting the chained CPI wouldn’t even bring Social Security’s deficit back to 2008 levels. Social Security needs fundamental reforms if it is to become both solvent and effective for those who need it the most. And politics matter as well: Conservatives need to pay as much attention to making government programs work better as they do to making those programs smaller.

If adopting the chained CPI for Social Security would be misguided, applying it to the income-tax code would be even worse. Most lawmakers are only dimly aware of how the CPI affects taxes, but it should tell them something that the Obama administration would not accept the chained CPI for Social Security without applying it to the tax code as well.

Currently, the dollar figures that define income-tax brackets, along with most credits and deductions, are indexed to the CPI-U. Since average incomes generally rise faster than the inflation rate, this indexing pushes income into higher tax brackets, even as tax credits and deductions become a smaller fraction of income. Over time, this “bracket creep” produces massive stealth tax increases — higher average tax rates bringing in more revenue for the government. The Congressional Budget Office projects that over the next 25 years, bracket creep generated by the CPI-U will raise personal-income-tax revenues by around 2.6 percentage points of GDP, a nearly one-third increase over the recent historical rate of 8.2 percent of GDP. Indexing the tax code using chained CPI instead would only exacerbate this autopilot trend toward higher taxes.

In other words, Republicans have already acquiesced in a one-third increase in personal-income-tax rates over the next quarter-century, and many appear eager to go for more. Yet Republicans would surely oppose such an increase if they understood it. Making matters worse, the largest rate increases will be on low- and middle-income households. The Congressional Joint Committee on Taxation projects that in 2021, 69 percent of the gains in revenue would come from taxpayers with incomes below $100,000, though they pay only 28 percent of total income taxes. Individuals in the highest income brackets would be left essentially untouched, because most of their income already falls into the highest bracket. Conservative reformers such as National Review’s Ramesh Ponnuru are pushing for a tax code that’s friendlier to families and middle-income earners. The chained CPI is hard to fit into that narrative.

Moreover, while the Social Security cuts due to chained CPI would stabilize at around 4 percent of outlays (being limited by the average recipient’s lifetime), the income-tax increases would keep growing in perpetuity. So while the chained-CPI debate focuses on spending cuts, over the long term the real action is on the revenue side. This is particularly ironic given that, in exchange for “accepting” the chained CPI, the White House is demanding tax increases in addition to those built into chained CPI.

A better approach would index income-tax brackets to the growth of incomes instead of prices, thus stabilizing tax revenues as a percentage of the economy. This would force Congress to make increase taxes explicitly whenever politicians wanted to increase revenues. If tax hikes are necessary, they should be debated and enacted openly, instead of raising taxes by stealth, as will happen with chained CPI.

It’s hard to see how chained CPI can be a win for conservatives. With congressional Democrats opposed, the narrative is already forming that President Obama only proposed using the chained CPI to appease congressional Republicans. But why should Republicans take the rap for a measure that weakens Social Security for the least well-off and institutes a large and regressive tax increase? Higher taxes and a less effective Social Security program — what’s not to dislike?

— Andrew G. Biggs is a resident scholar at the American Enterprise Institute and former principal deputy commissioner of the Social Security Administration.

Andrew G. Biggs is a senior fellow at the American Enterprise Institute. He previously served as the principal deputy commissioner of the Social Security Administration, as well as working on Social Security reform for the White House National Economic Council in 2005.


The Latest