Private payrolls declined in September by 32,000 in key ADP report coming amid shutdown data blackout

[Wiz Imp]

15. Nice try. But you failed big time. You'll find that I've never typed a single word in any post here

that contradicts a single word that Erica Goshen said.

Not a single person here has ever claimed that the data quality would go down because of falling response rate (that I've seen. If someone did, I certainly never disagreed with them). Every single person posting here who doesn't want to believe that data released under a Trump administration could ever be honest, have insisted Trump (his administration) is intentionally manipulating the data. I've been trying to show people for months that they were wrong - the data from BLS to this point has been completely honest. Erica Goshen in fact effectively echoes my words about Acting Commissioner Bill Wiatrowski.

The acting commissioner, Bill Wiatrowski, is a career civil servant, incredibly dedicated and a skilled leader. So, I don’t think [Trump’s actions] will affect the way the two surveys are conducted and how the data are processed unless there is an outside attempt to interfere with its operations.

The rest of her statement is exactly what I would have said, if anyone had ever argued about the quality of data due to falling response rate. But that never happened. Everybody constantly says the data are manipulated or cooked or whatever other word they want to use for dishonest. Every one of these people has effectively argued that the Trump administration is making up all the economic data they are releasing. They are provably wrong and Erica Goshen agrees with that.

Of course Erica Goshen knows what she is talking about. Any statistician could tell you the lower the response rate to a survey with a stratified sample, the less reliable the data is going to be. For most of my career, you would have heard me complaining about low response rates. And as much as they can impact national data, the impact on State and Metropolitan Area data which I dealt with is significantly greater due to the samples being much smaller for the smaller geographic areas.

Could Trump's comments cause the response rates to go even lower? Of course. But that is not manipulating the data. It's just a natural side effect from anti-Government rhetoric.

If people want to complain about the quality of the data due to low response rates or Trump's rhetoric possibly lowering response rates which in turn could lower data quality, you won't find me disagreeing with them (unless they make incorrect assumptions about how the surveys work). But I have no patience for people who insist the data is being intentionally manipulated. It is simply not true and I will continue to call it out when I see it.