“About 20% of the population over this whole entire period get exactly a 3% wage increase every year. Exactly. Not 2.94, not 3.02, exactly 3%. … Firms give most of their workers 3% most of the time.”

Literally on your paycheck. You know, University of Chicago uses Workday. That is another competitor. Workday and ADP are the two, you know, of the most prominent of the payroll processors in the country. And so why is that important is when I want to try to measure your wage, your exact wage, the way we do it through surveys is I would say, Hey, Tess, how much did you make last year? And you give me an estimate. And then I say, Hey, how many hours did you work last year? And you give me an estimate. And then I take the noisy measure of your earnings and the noisy measure of your hours, and I make a wage rate. And then I do the same thing for you tomorrow when I survey you tomorrow. And then when I take what happens to you, the change in your wage, it's a lot of noise because the level has a little bit of noise in it. And anytime I ask you, you know, about to try to approximate your wage by giving me your income and your earnings, there's a just measurement error. And changes, that measurement error gets much bigger. And so things, other data sets, survey data sets, are not really good to measure how wages adjust. And so the administrative payroll data, I know exactly how much Tess is making. And so as a result, I don't have to guess what your earnings or your hours are. You don't have to approximate. I get an administrative record. And so seeing that allows me to measure precisely how people's wages are changing. One thing you could get with this that you can't get with other surveys is just the fact that 3%, about 20% of the population over this whole entire period, get exactly a 3% wage increase every year. Exactly. Not 2.94, not 3.02, exactly 3%. And then if you include the people approximately 3%, the number is much higher. But once we go to the, you know, the survey data, if you rush estimated your hours and earnings, your hours and earnings in each year, I'm going to get a lot of noise in your change in your wage over periods of time. The other thing about the payroll data that is immensely useful for us is I could measure every worker in the same firm. So I could see how firms are setting wages. When I go to household surveys, I might get you, but I don't have all your coworkers. And so here, because the company contracts with ADP, if one worker's wages are processed by ADP, every worker is processed by ADP because they process for all the workers. And so here I could see how wage setting looks like within a firm. How many people are exactly getting 3%? Are some of us getting 3% and some of us getting 10%? Some of us getting 1%. I could then see the distribution of wage changes to see how firms are changing their wages. And what you see, firms give most of their workers 3% most of the time.