Report | Wages, Incomes, and Wealth

Securing the wage floor: Indexing would maintain minimum wage value, provide predictability to employers

Briefing Paper #177

Download PDF


In 1938 Congress enacted the federal minimum wage, originally setting it at 25 cents per hour, as part of the Fair Labor Standards Act (FLSA). From the beginning, it was clear that the minimum wage would have to be raised periodically to keep up with rising prices and wage levels. Original proposals for the FLSA provided for a commission that would set the minimum wage after a public hearing and consideration of cost-of-living estimates provided by the Bureau of Labor Statistics (BLS). By this procedure, the wage would have been updated according to changes in the standard of living and inflation. The version of the FLSA that became law, however, left action on future increases to Congress and the president.

This inherently political system has, not surprisingly, led to inconsistency in maintaining the value of the minimum wage. Depending on the dominant political philosophy of the times, the minimum wage has trended up or down in its inflation-adjusted value. The trend since its peak in 1968 has been downward, and the current wage of $5.15 is the lowest in 50 years (see Figure A).

Figure A

The long-run trends in the value of the minimum wage reflect a repeated pattern of declines, as inflation has taken its bite, followed by sharp hikes by Congress and the president after several years that may or may not restore the wage’s buying power. Before 1968, most increases more than made up for the decline in value following the prior increase—each boost bringing the minimum wage to new highs relative to prices. Since 1968 the increases have often not made up for the prior decline.

Many of the declines in the minimum wage, as well as the spasmodic boosts by Congress and the president, have been substantial. The largest decline in value relative to inflation was 34% between 1978 and 1989—a decline of $2.74 per hour in 2006 dollars (or $5,700 a year for a full-time, year-round worker). Since 1938, the minimum wage has been raised 10 times (counting phased-in, multiyear increases as a single increase), and most increases have fallen in the 10-20% range. The current period is the second longest without a hike: the value of the minimum wage has dropped by 20% since the last increase in 1997.

The up and downs of the minimum wage relative to prices or purchasing power has important implications. A minimum wage that doesn’t buy as much as it used to is a minimum wage that is placing a lower value on work and doing less for minimum wage workers. But another important principle of the minimum wage is that the gap between the standard of living of low-wage workers and the standard of living of other wage earners should not become excessive. By this measure as well, the minimum wage has declined and varied substantially. During the 1950s and 1960s the minimum wage varied between 44% and 56% of the average wage.1 It is now only 31% of the average wage (see Figure B).

Figure B

Such a variable, inconsistent, and low minimum wage is literally “no way to run a business.” The gradual declines and abrupt increases take their toll both on workers who rely on the minimum wage and see it fall over time, and also on businesses, which find the wage costs for their low-wage workers to be unpredictable and subject to the political winds.

This is why the idea of putting in law automatic annual adjustments, or indexing, has gained increasing support. Other federal and state laws routinely use indexing to ensure that benefits and other dollar values do not lose their value to rising prices. Indexing the minimum wage is a common policy in many other countries and is becoming increasingly common among the U.S. states that have their own minimum wage laws. Indexing the minimum wage assures that low-wage workers do not lose ground to rising prices—ground they can ill-afford to lose—and offers their employers predictable and steady changes in the legal standard.

Choosing an index

Automatic annual adjustments are a sensible solution to this long-standing problem with the minimum wage. But the question remains, how should the size of those adjustments be determined? Most other developed countries either have implemented automatic increases based on rising prices or require regular meetings of boards authorized to increase the minimum wage based on several factors—usually including rising prices and in some cases rising wages and economic growth. In the Netherlands, automatic increases are based on average wage growth (the increases can be suspended under special circumstances). U.S. states that have indexed their minimum wages have used price indexes to tie the wage to inflation. This form of indexing protects low-wage workers from declines in their standard of living due to the eroding real value of their minimum wage. Another form of indexing could link the minimum wage to average or median wage levels.

Linking to inflation measures 
States that have indexed to prices have used forms of the BLS’s consumer price index (CPI), which is used for indexing a wide variety of federal and state benefits, exemption levels in the tax law, and pension benefits; it is also widely used in private sector contracts. There are a number of versions of the CPI to choose from, each widely used for government or private sector indexing.

  • Consumer Price Index—All Urban Consumers (CPI-U). The title for this index is somewhat misleading in its use of the word “urban,” since the index covers 87% of the U.S. population. The only populations excluded are those serving in the military, the institutionalized population, and those living in isolated locations. At least a portion of every state is included. Because it would be impractical for the BLS to collect prices from every transaction within the covered areas, it relies on a carefully constructed sample of 87 locations in determining price changes. The sample has been selected to be representative of the entire population covered by the CPI-U.
  • Consumer Price Index—Urban Wage Earners and Clerical Workers (CPI-W). This subset of the CPI-U, covering employed wage earners and clerical workers, is designed to reflect the changes in prices faced by working Americans. In practice, it tracks the CPI-U closely. It is, notably, the index used to adjust Social Security benefits.
  • Consumer Price Index for Specific Areas. The BLS others versions of the CPI-U and CPI-W to cover four regions of the country (Northeast, Midwest, West, South) and 27 metropolitan areas.


There are merits to each of these as an index for the minimum wage. The CPI-U is the most widely known of the indexes, the CPI-W is most closely targeted at the costs faced by workers, and the area indexes are useful indicators for the areas they cover. The choice for most states has been based on which indexes they use in their other laws.2

Linking to average wages&nbs

Linking the minimum wage to average or median wages, instead of prices, could lead to significantly different adjustments to the minimum wage. Over the past 50 years the CPI-U has gone up 740% (i.e., prices have gone up by a factor of 7.40), while the average wage for production, non-supervisory workers (a good proxy for the general workforce) has increased by 922% (i.e., wages have gone up by a factor of 9.22). Restoring the minimum wage to half the average wage, the approximate level of the 1950s and 1960s, would result in a substantial immediate boost in the minimum wage—to $8.40 per hour. It would also help stem rising wage and income inequality and connect the pay of low-wage workers more closely both to what employers can afford and also to rising living standards in a growing economy.

The indexing debate

The indexing of the minimum wage has been challenged on the grounds that it results in minimum wage levels that are too high or too low. Among the arguments:

  • Indexing the minimum wage from a low level will keep it at a low level, valuing work too little and leaving low-wage workers unable to support themselves and their families.
  • During periods of economic downturn or high inflation, price-based indexing will lead to a runaway minimum wage that causes economic harm.


Does indexing keep the minimum wage low? 
If indexing is put in place at a time when the minimum wage is at a historically low level, will substantial, needed in-creases become politically difficult to obtain in the future? It is important to remember that indexing a minimum wage does not increase the minimum wage relative to prices or average wages; rather, it locks the wage in place. For example, if the minimum wage is $6.00 per hour and prices go up 3%, a price-indexed minimum wage would go up by 3% to $6.18 per hour. That $6.18 would, however, on average, buy the same amount as $6.00 did the year before. So, in terms of purchasing power and quality of life the minimum wage has stayed the same. If the minimum wage is too low from the start, indexing will maintain its limited buying power and perhaps stand in the way of needed increases.

This concern is not based on a fundamental characteristic of indexing but instead on a reading of the politics of the minimum wage. The fear is that once indexing is put in place it will take the political steam out of proposals for future increases. If that’s true, indexing from too low a level will, indeed, lock the minimum wage at too low a level. The validity of this concern is difficult to judge at this point. The states that have thus far indexed their minimum wages for inflation have either started at relatively high minimum wage levels or only begun indexing recently.

Could indexing cause the minimum wage to spiral too high? 
Minimum wage opponents have raised the alarm that in times of economic downturn, or high inflation, price-based indexing will cause the minimum wage to reach excessive levels that will cause job loss and further drive up inflation. There are two bases for this concern. The first is that a minimum wage indexed to inflation would increase, in nominal terms, at too rapid a rate during a period of high inflation. Second, during an economic downturn, average wage growth can lag price increases. A minimum wage rising significantly faster than other wages could distort labor markets—making it harder for employers to employ low-wage workers.

But minimum wage increases due to price indexing, relative to the increases that have been enacted legislatively, are small. With the exception of one, highly unusual period discussed below, inflation rarely exceeds 5% a year; it has exceeded 4% only once in the last 20 years (4.1% in 1998). Given that recessions rarely last more than a year and rarely coincide exactly with the effective date of an indexed increase, the overlap between a higher minimum wage and a period of recession is likely to be small (and, as explained below, indexing will not lead to excessive minimum wages leading into recessions). A 5% increase on the current $5.15 would bring the wage to $5.40 per hour (which would still be its lowest real value in 50 years). Such an increase would add less than 1/100th of a percent to the national wage bill. Thus, during recessions, indexing will typically cause only a single, small, nominal minimum wage increase that will be effective only for a portion of the recession period.

Moreover, this phenomenon would not be a new one: the current system of federal minimum wage adjustments also results in increases during recession years. The federal minimum wage was enacted during the Great Depression, and was increased again before its end, in part to preserve wages in the face of extreme unemployment. Since then, four increases have gone into effect during recessions and four more have gone into effect during the same years as recessions, but shortly before or after their beginning or end. Thus, increases during periods of economic weakness have not been seen heretofore as economically damaging.

A measure of whether growth in a minimum wage is excessive is to compare it to average wage growth. A minimum wage growing consistently and substantially faster than average wages might distort the labor market. This, however, proves not to be a concern with respect to indexing. An indexed minimum wage might, during some periods of high inflation or economic sluggishness, grow somewhat faster than average wages, but in most other periods it will lag average wage growth. As noted above, if the minimum wage had been indexed to the CPI-U over the last 50 years, it would have increased in nominal dollar value by 740% over that period. Over that same 50 years average wages grew in nominal terms by 922%.

Figure C shows the change in the nominal average wage versus the change in the minimum wage had the minimum wage been indexed to the CPI-U in 1956. Under this scenario the minimum wage would have occasionally gained ground on the average wage, but never substantially. Indexing for prices would not have caused the minimum wage to reach excessive levels during times when the economy was doing poorly or inflation was high—the lag in the more typical years prevents that. One reason for this is that, during periods of high inflation, average wages grow at a faster rate than during periods of low inflation, so a minimum wage adjusted for inflation doesn’t necessarily gain much ground against average wages. For example, consider the 10 years from 1973 to 1982, the period of highest inflation in 50 years. Prices rose between 5.8% and 13.5% a year, but eight of those 10 years also rank in the top 10 for nominal average wage growth. In all of those years, average wages grew at least 5.8%, and in three of those years wage growth exceeded inflation.

Figure C

During periods of high inflation prices generally rise across the board. Wages go up faster, inputs purchased by businesses are on the rise, and the prices businesses charge and revenues they receive increase. The cost of indexing the wages of low-wage employees is a very small part of this picture. Minimum wage workers should not, uniquely among workers, get no increases in their wages during periods of high inflation. They are the workers most in need of having their incomes keep up with rising prices.

Indexing in
the states has been a success

Four states—Washington, Oregon, Vermont, and Florida—now index their minimum wages to prices using the CPI, and in these states wages have maintained their purchasing power for hundreds of thousands of workers without adverse effects. Washington indexed its wage through the most recent recession with no ill-effects; during that period, low-wage sectors of the economy that are affected by the minimum wage actually outperformed higher-wage sectors.3 Nor have the annual inflation adjustments in Washington and Oregon (the two states with the longest history of indexing) fueled inflation.4 BLS price indices for areas within Washington and Oregon have actually dropped relative to the national inflation rate after the states began increasing their minimum wages.5

The minimum wage can have little impact on inflation 
It is not surprising that the minimum wage has not fueled inflation in the states where it has been tried. Although the minimum wage is important as a demonstration of commitment to the value of work and as a support for low-wage workers, there are simply not enough low-wage workers with enough aggregate income to substantially affect overall prices. A national minimum wage increase from $5.15 to $7.25, a 41% boost—a much more substantial hike than any increase that would result from indexing—would raise the total labor costs of the country by only 0.22%, too little to cause a noticeable increase in inflation in the context of an economy experiencing constant adjustments in wages and prices on much more substantial scales.


The minimum wage has, over the years, had many ups and downs. Breaking the cycle of declining purchasing power and sudden increases would benefit low-wage workers and their employers. Ensuring that minimum wages keep up with either the cost of living or average wages is consistent with a thriving economy. The increases that are seen with minimum wage indexing are moderate, by several measures, and pose no risk to the economy. Indexing the minimum wage is a modest step forward in ensuring that low-wage workers do not see their wages decline over time and that low-wage employers can anticipate predictable changes in the minimum wage.

—October 2006



1. The average wage used here is for production, non-supervisory workers.

2. Among states that currently have indexing, Washington and Florida use the CPI-W and Oregon and Vermont use the CPI-U.

3. See the Economic Opportunity Institute’s “Still Working Well: Washington’s Minimum Wage and the Beginnings of Economic Recovery” (January 2004, available at; and Jeff Chapman’s “Employment and the Minimum Wage: Evidence From Recent State Labor Market Trends” (May 2004, the Economic Policy Institute, available at

4. Although the Washington minimum wage is the highest in the country, it is not because of indexing. Washington started with the highest minimum wage, and indexing has kept the state in the lead.

5. The Seattle-Takoma-Bremerton rate was between 1.2 and 1.4 percentage points above the national rate in the two years prior to the beginning of the Washington minimum wage hikes. Since then it has been between 1.5 points below and 0.8 points above the national rate. The Portland-Salem index was 0.3 and 1.1 points above the national rate in the two years preceding the major increases in the Oregon minimum wage and has been between 0.9 points below and 1.1 points above the national rate since then. There is no evidence that the minimum wage is having any impact on overall inflation in these two states.

See related work on Minimum Wage | Wages, Incomes, and Wealth

See more work by Michael Ettlinger