One of the more contentious debates about the changes in the U.S. economy is the question of whether wages have grown in tandem with productivity growth. The debate hinges on the particular merits of different tweaks to data. But one thing that almost every participant in the debate agrees upon is that analysts should be looking at total compensation—not just wages– when trying to figure out how productivity turns into income gains for workers. Yet, changes in compensation also affect other trends, namely economic inequality, which in turn may well link back to who gets the fruit of productivity growth and rising living standards.

Over the past 40 years or so, employers have increasingly compensated workers not just in the form of wages and salary, but also other benefits such as health insurance or retirement plans. Looking at wage and salary compensation alone doesn’t give the total picture of how much workers are getting paid if, for example, employer-subsidized health insurance costs aren’t factored into the mix.

Wages were a larger share of total compensation back in 1970s than they are today, but the present-day gap between wages and compensation also varies across the income ladder. A recently released paper by Kristen Monaco and Brooks Pierce, research economists at the U.S. Bureau of Labor Statistics, looks at the changes in inflation-adjusted wages and compensation from 2007 to 2014 using data from the National Compensation Survey. Looking just at wages, they find a checkmark-shaped pattern in wage growth. Wage growth was weakest for workers near the middle of the income spectrum (the median fell by 4 percent), slightly higher for those at the bottom, and highest for those at the top.

When Monaco and Pierce then looked at total compensation growth, they again find the checkmark shape. But placing the two curves over each other reveals an interesting trend. Growth in total compensation for lower-paid workers was slower than wage growth in that same spot on the wage spectrum. The exact opposite happens for highly-compensation workers: Compensation growth is faster than wage growth. But wage growth and compensation growth don’t vary that much for those in the middle.

What this means is that compensation inequality grew more than wage inequality did between 2007 and 2014. This study only looks at trends since the Great Recession’s onset, but there’s evidence that compensation inequality has grown faster than wage inequality since the 1980s as well.

Now let’s turn to the relationship between total labor compensation and productivity. Recent analysis on this question by Harvard University economist Robert Lawrence for the Peterson Institute for International Economics tracks average productivity to average compensation instead of average wages. If the question you want to answer is ultimately how well labor as a whole is being compensated for its productivity, this data choice is the right way to go. Lawrence finds evidence of a break between productivity and average labor compensation around 2000, just as he did in a recent National Bureau of Economic Research working paper. This break means that labor as a whole is receiving a declining share of income, with the gains from productivity go to the owners of capital instead.

But what if we want to think about how productivity growth results in rising standards of living for workers on different rungs of the economic ladder? Then we need to look at what’s happening to the distribution of compensation. It may have been that compensation for labor as a whole tracked productivity until 2000, but it’s not clear how well productivity growth was translating into growth and higher living standards for all workers.