CEA has released a "primer" on wage measurement. Much of the substance is reasonable (with one exception noted below). But it does almost nothing to change our understanding of wage growth. THREAD whitehouse.gov/wp-content/upl…
CEA focuses on four changes all of which raise the rate of real wage growth in the present but also raise it by a similar amount in the past and leave the overall pattern unchanged. Let me discuss them in turn.
FIRST, COMPOSITION EFFECTS. Headline wage growth measure suffers from a composition effect that artificially reduces wage growth as younger, lower-paid workers enter and aging affects the share of the workforce in peak earning years.
This is correct and why it is worth looking at the Atlanta Wage tracker which addresses this issue. frbatlanta.org/chcs/wage-grow…
CEA finds that adjusting for this composition effect adds 0.3 pp to the annual growth rate. But sometimes it changes the pattern. For example, in 2009 and 2010 reported wage growth was artificially boosted when low-wage workers were fired and now the reverse is happening.
SECOND, INCLUDING BENEFITS. CEA is correct that it is, in theory, desirable to include benefits like health insurance, bonuses, paid leave, and the like. The ECI includes these while the standard Average Hourly Earnings measure does not.
You could debate whether workers value benefits as much as they value wages. But absent certain tax distortions, unless companies are making systematic errors they are probably not giving workers stuff that costs the companies more than the benefits to the workers.
As an aside, much of the progressive advocacy for mandated or publicly provided benefits like sick leave implicitly assumes that these are worth more to workers than the equivalent amount of cash.
THIRD, INFLATION ADJUSTMENT. Most economists use the CPI but I agree with CEA that there is a case for using another inflation measure, the PCE price index, which I sometimes use myself. The PCE price index is chain-weighted so it does not suffer from substitution bias.
The advantage of the CPI is that it uses a market basket that more closely approximates what consumers are buying with their wages. (The PCE also includes what is purchased by Medicare, nonprofits and the like).
Probably the ideal would be to use the chained CPI which corrects for substitution bias but has the correct basket. This would put real wage growth between the CPI version that is often reported and the PCE version CEA prefers.
None of these inflation measures fully account for quality improvements so all probably show too much inflation and too little real wage growth. This bias is likely lower today than in the past. brookings.edu/wp-content/upl…
FOURTH, CEA GOES WRONG IN ADJUSTING FOR TAXES. It is reasonable to look at INCOMES before and after taxes. But highly unusual to apply this to wages. Wages are meant to be a measure of labor market outcomes not an overall indicator of well-being.
In 2014, for example, the ACA subsidies and Medicaid expansion increased after-tax incomes broadly construed. But it would have been silly to add the subsidies to wages and describe it as wage growth.
In addition, you need to take into account how the tax cuts were distributed (more of them were for the top) and how they will be repaid. People get more money but also are on the hook for repaying more public debt.
FINALLY, CEA does not address the issues of tradeoffs and noise in data. Sometimes we look at more timely data (like AHE) even though it is not perfect, in part because we think the imperfections don’t change the story.
Also, all data is noisy so best to combine multiple data sources. The Goldman Sachs wage tracker and Jared Bernstein wage tracker, for example, use principal components to form a statistically-based weighted average of different series to exclude noise. washingtonpost.com/news/postevery… 
And it would be nice if CEA did not just make these adjustments in talking about the economy since January 2017 but applied them earlier. I’m thinking about the opening of this report, for example, which did not make these adjustments.
IN CONCLUSION, most of CEA's points are not new but many of them are under-appreciated so good to call attn to them. But they don't change our understanding of wage growth in the economy (beyond rebutting silly statements you sometimes hear about no wage growth since the 70s).
Oh, and here is my explanation for slow wage growth. END THREAD vox.com/the-big-idea/2…
