What happens when statistics get politicized?

From Marketplace, I'm Sabrina Beneshore in for David Brancaccio.

On Friday, President Trump fired the top Bureau of Labor Statistics official after the Bureau

issued a weak jobs report.

The report had also revised earlier estimates of job creation in May and June down by 258,000.

The president claimed the numbers were being manipulated.

Julia Cornado is founder and president of macro policy perspectives and a professor at UT Austin and is here to talk about it.

Hi, Julia.

Good morning.

Is there any evidence that, as President Trump claims, the job market numbers were being manipulated for political purposes?

Absolutely not. The Bureau of Labor Statistics is staffed by excellent dedicated public servants,

trying to measure the U.S. economy in real time. Re revisions are not unusual. There is no evidence of that.

Over the weekend, National Economic Council Director Kevin Hassett changed the message of it and told NBC News that,

well, it was actually the magnitude of the revisions that justified the firing. Do such major

revisions suggest some kind of lack of competence or wrongdoing?

No. In fact, there is an economic literature that documents how the revisions to economic data do

become larger at turning points in the economy. That is, when the economy is shifting gears, let's say

downward, then early estimates don't tend to capture that as well, and you tend to get larger

revisions. Does the fact that the president fired the commissioner of the Bureau of Labor

Statistics because he didn't like the numbers, erode the trust in

or integrity of those numbers, especially if he's going to replace that person with, you know,

someone who prefers in the next few days. It definitely erodes trust. We can't assume that the

...