From The Hill:

If you’re looking at the CBO for accuracy, you’re looking in the wrong place,” White House press secretary Sean Spicer said Wednesday.

Since Mr. Spicer’s BA is in government rather than economics, I thought it useful to consult some documentation based on quantitative assessment. Here are three graphs from CBO, “CBO’s Economic Forecasting Record: 2015 Update,” February 2015.

Source: CBO, “CBO’s Economic Forecasting Record: 2015 Update,” February 2015.

here is a summary for three key indicators, CBO vs. OMB vs. Blue Chip:

Source: Excerpt from Summary Figure 1, in CBO, “CBO’s Economic Forecasting Record: 2015 Update,” February 2015.

Source: Excerpt from Summary Figure 2, in CBO, “CBO’s Economic Forecasting Record: 2015 Update,” February 2015.

Overall, mean bias in CBO’s two year projections is pretty small for GDP, nominal GDP growth, and inflation. Root mean squared errors (RMSE’s) are comparable to Blue Chip forecasts (and a little smaller than OMB), which is pretty remarkable given (1) CBO projections are conditioned on current law, and (2) Blue Chip forecasts are the averages of individual forecasts, and averages of forecasts are typically more accurate than single forecasts.

I’m not sure who Mr. Spicer believes would provide better forecasts. Since, as in many other instances, he has not provided additional information, we are in the dark in this regard.

However, we can be fairly sure with the directive to OMB and CEA to reverse engineer Administration forecasts to hit 3-3.5% growth [1], OMB growth forecasts will eventually have much bigger RMSE’s than CBO’s.