The fact remains that bureaucracies receive their marching orders from the exec branch and the methods which are used to calculate statistics are changed periodically.
Sort of. The executive branch gives the budgets, and can direct broad program or admin changes, but not operational matters. And major changes to statistical methods are suggested by committees and take years to implement.
Neither the President nor anyone in his administration can directly influence the reports.
Again I don't particularly have a problem with BLS personnel, but if you think they are free to calculate statistics of their own volition I think you're kidding yourself.
I'm glad you agree that BLS personnel stick by established guidelines and procedures and cannot just change things or order subordinates to change things. There is a lot of supervision and oversight. And even when/if someone does try to make their own changes, as was the case of a fired Census worker a few years ago, there is usually no effect.
It makes me think of how politicians use the GAO and CBO to calculate the effects of proposed legislation. What people don't realize is that all the criteria are supplied by the politician. "What will the outcome of my legislation be IF inflation remains x, the cost of resource y remains z". These groups are often just calculators using outside supplied criteria.
Statistical agencies cannot do that. BLS employees are specifically forbidden from what if's and BLS cannot comment on the effect of policy.
The most glaring example of changing methods is the inflation rate. The method of calculation has changed drastically over the years. What are the chances they are changed to make the rates less attractive to the electorate?
Zero. Nor were they changed to make the rates more attractive. They were changed in the 1980's and the 2000's because it was clear that the previous methods overestimated inflation. It's not even an argument that a Laspeyres index gives the upper bound of change..it's a plain fact. There is also no question that hedonic adjustment of quality change (when direct adjustment is not possible) is much more accurate than excluding series that have quality change or imputing the change.
Some specifics and details are of course arguable, and the Current Expenditure Survey, which is the source of weights for the CPI is being heavily revised due to known collection and data issues.
Whether we like it or not bureaucracies are political entities and not taking their pronouncements with a grain of salt is foolish.
But rejecting them outright is also foolish.
The original points of the end of the verizon strike. summer hiring and future adjustments(usually down) are all valid points.
Add to that calculating who is in the workforce calls for perhaps more than "grain" of salt.
Those on strike are considered employed for Unemployment rate calculations, and may or may not be considered employed for the official jobs numbers depending on the length and timing of the strike. Summer hiring is adjusted for by seasonal adjustment.
What specifically do you think the issues are with calculating the labor force level or rate?[/quote]