These days, wages in the United States are doing something extraordinary: They’re growing faster at the bottom than at the top. In fact, recent growth for workers with low wages has outpaced that for high-wage workers by the widest margin in at least 20 years.
The main story here is the long economic recovery, now entering its 11th year. For much of the early phase of this recovery, wage growth for the bottom group was weaker than for others, but it began gradually accelerating in 2014 as unemployment continued to fall. This was around the same time the labor market started tapping into people some economists had all but given up on as workforce participants, such as those who had been citing health reasons or disability for not having a job.
But there has been another factor at play: the rise in state and local minimum wages.
For the last decade, the federal minimum wage has been unchanged at $7.25 an hour. But over that period, dozens of states and localities have enacted their own minimum wages or raised existing ones. As a result, the effective U.S. minimum wage is closer to $12 an hour, most likely the highest in U.S. history even after adjusting for inflation.
And with two dozen states and four dozen localities set to raise their minimums further in 2020, the effective minimum wage will keep rising this year.
These state and local actions are affecting wage data, especially for workers at the bottom. To get a sense of this impact, I have used data in the Current Population Survey to look at minimum wage workers as a group and calculate the pressure their wage gains have put on aggregate wage growth over time, controlling for compositional changes in the share of minimum wage work.
Note that this approach doesn’t settle whether minimum wage increases are a net benefit to Americans, since among other things wage data will by definition capture only those who stayed employed after an increase. If people were laid off because of a minimum wage increase, their loss of wages wouldn’t factor into the average.
This analysis shows that growth in average wages has been running about 3.9% per year in the Current Population Survey over the past two years, a bit firmer than the pace right before the Great Recession but below the near 5% reached in 2000.
But increases to minimum wages at the state and local level have put 0.4 percentage points of upward pressure on this recent growth. Absent that pressure, wage growth in the Current Population Survey over the last two years would have been 3.5%. That’s still a fine result, but it’s a bit cooler than the unadjusted data suggest.
Wage pressure from minimum wage workers is magnified when you look at only the lowest wages. That’s because while minimum wage work makes up about 6% of all usual hours worked, it’s around 13% of hours worked by Americans in the bottom third of wages.
As the analysis has shown us, wage growth at the bottom is doing well. It has been around 4.1% over the last two years — above the 3.6% at the top end and above the overall average of 3.9%.
But absent the pressure from minimum wage workers, growth at the bottom would have been closer to 3.3%.
It’s important to keep the effect of these minimum wage increases in perspective. The increases aren’t responsible for most of the wage growth, or for most of the acceleration in wage growth, during this recovery. Even among the bottom third, minimum wage workers have contributed around a fifth to a quarter of wage growth over the last two years.
As notable as the recent rise in state and local minimum wages has been to this effect, it has probably not been as important as the tightening labor market. In a tight labor market, firms have to compete more to hire and retain the workers they need, which among other things gives those workers more bargaining power to bid up their wages.
The Federal Reserve chair, Jerome Powell, has argued that reaching workers traditionally “left behind” is one of the most compelling reasons to sustain the expansion for as long as possible.
Still, this analysis suggests these minimum wage increases are having a meaningful impact on wages, at least for the employed workers who benefit from them. For the bottom third, state and local minimum wage increases have probably been the difference between the wage growth just before the economic crisis and the wage growth that is now above that pace.
But that benefit also brings with it a cautionary note for policymakers.
Economists look to wages as one thermometer of how hot the economy is getting: Accelerating wages may eventually spill over into higher prices and signal an economy at capacity, although this hasn’t happened yet in this recovery.
But these continuing increases to state and local minimum wages — and any possible future action at the federal level — could skew wage data, making the American labor market look tighter than it actually is.
The recent rise in minimum wages, although not producing a giant effect, still might suggest overall wage growth is progressing about a year further along than the reality. For low-wage workers, whom policymakers are citing more often, the minimum wage effect can be worth closer to two years’ worth of wage acceleration.
The risk of misdiagnosing an overheating economy is one reason it’s important to be clear and precise about what role minimum wage increases have played in recent wage growth: They have been important, but they’re most likely not the biggest factor.
Methodology: This analysis defines “minimum wage pressure” as the growth in the effective minimum wage — the average binding federal, state or local minimum wage received per hour of minimum wage work — over 12 months, weighted by the share of minimum wage work at the beginning of the 12-month period. In a shift-share framework, this is the equivalent of the compositionally adjusted contribution to aggregate wage growth from minimum wage workers. The analysis uses average hourly wages for private nonfarm wage and salary workers calculated from the Current Population Survey Outgoing Rotation Group. It takes hourly wages as given when available for hourly workers; for others, it divides usual weekly earnings by usual weekly hours. The analysis imputes usual hours when unavailable or varying, and adjusts weekly earnings for top-coding using a log-linear distributional assumption. It also trims outlier hourly wages. Despite these adjustments, wages in the CPS invariably differ from average hourly earnings reported in the Current Employment Statistics, another Bureau of Labor Statistics data source for wages, because of differences in scope, design and concept. The analysis uses the same methodology for calculating state and local minimum wages, and for identifying minimum wage workers, as in a previous analysis from April 2019.