It seems unnecessary, given that even when it spends money to research an issue, and doesn’t bother to hide the results, it will still blatantly ignore the recommendations in favor of whatever it wanted to do in the first place.****
How to calculate or measure different types of data is an ongoing discussion on any fields which obtain and employ such data; how to compare data obtained with different methodologies (and in fact, how to prove whether two methodologies are equivalent) is a branch in any such field. Often, cherry picking begins by picking one methodology over another, since the biases of different methods are known (all measurement methods introduce some deviation from reality, but some more than others and in different directions). Waiting times for medical systems; unemployment percentiles; program viewership; the color of samples. When governments are involved, those discussions tend to be pretty public; whether the public and the politicians involved can understand the discussion or not is a different matter, but the information is available. Getting the decision-makers to understand what they’re looking at is as much of a problem in private enterprises as in governments: as big and as small.
This is why scientists start with a question instead of a statistic. These are two different questions with two different answers:
[ul]
[li]How many people that are trying to find a job are not working?[/li][li]How many people that could be working are not working?[/li][/ul]
There are also lots of followup questions on the differences between those answers. So then the headline is “Unemployment at 4.9%” and people start saying, “but, but, but…” because they’re starting with a number, not a question.
Even among experts in a field, one person’s cherry picking is another’s quality control. If they’re both honestly trying to find the best answer, then there’s a good chance both arguments have merits. If one or both of them is trying find a particular answer, then we’re possibly into fraud.
That of course is not fabricated statistics, it is simply PR declarations with made up numbers by your President.
It is not a more honest statistic, it is a different statistic
No, not in any way.
It is making the mistaken assumptoin that all who are not in the workforce are interested in a paid position - which is far from the truth for many reasons - such as women who do not desire to be in the labor force, and confuses other categories of non work for the demand for work.
The seeking employment statistic is a key statistic for having an estimation of the pressure in the labor market, the active supply versus the active demand.
That is the economics.
Other issues are part of certain left political disputes relative to policy, which may be correctly based or not, but are a different question from the economic demande question.
Everyone knows 43.27% of statistics are just made up.
You speak as if The Government was some evil monolith. It isn’t either one. But if you are a foxnews junkie or something, you might believe a stupid lie like that.
dasmoocher writes:
> I thought everyone knew that 86.894% of all statistics were just fabricated.
Bryan Ekers writes:
> Everyone knows 43.27% of statistics are just made up.
O.K., this is not acceptable. We’re going to have to determine which is the correct number. Luckily I am about to start a foundation to work out the true percentage. It’s going to take years to work out and will employ a lot of scientists to do research. I will be the head of the foundation, of course. I expect everyone contributing to this thread (or, indeed, everyone who’s ever posted to the SDMB) to make a large charitable contribution to the foundation. I will only ask for a mere $500,000 per year to run the foundation.
Well, when seasonally adjusted, they both give you 110% which coincidentally, as you would be aware, is the standard metric of measurement for sportsperson effort per event, divided by the number you first though of.
I used to work at NIST and this xkcd cartoonwas taped to the breakroom refrigerator.
We also had a saying: “The person with one watch knows what time it is; the person with two watches is never sure.”
Honest discussion of workforce participation try to remove people like me, retired and not interested in finding work, to find discouraged workers who don’t show up in the regular unemployment figures but who still are a problem.
The bad odor in the US comes from certain politicians who use workforce participation when out of office to claim that there is a major unemployment problem and then use regular unemployment when in office to show how much better they’ve made things.
Is there an official breakdown of the participation numbers beyond actively looking for work and not? I’ve never seen one quoted, and I’d think it would be hard to determine.
Okay, while I can appreciate your reluctance to link to the America-hating fuckstick’s Twitter feed, I would even MORE appreciate a cite to establish that your quote actually came from him (don’t take if personally. Smapti has demonstrated that his tweeting voice is reproducible).
You might find the U4 and U5 statistics useful.
This unfolding story about Indian Prime Minister Modhi attempting to suppress statistics that make him look less than totally brilliant and effective is a useful illustration of the competing intention of politicians and number-crunchers, and also the imperfect ability of the former to interfere with the latter.
Ask and ye shall receive: Trump’s twitter feed, Jan. 27
Generally numbers from the natural sciences are very accurate. However, I absolutely don’t trust things like labor statistics, as they are often very politically driven. They generally won’t report an incorrect wrong number (unemployment at 6% when their numbers say 8%), but will gladly collect data in a way that produces a result that is not accurate and fits a political agenda.
For example, numbers from the Bureau of Labor Statistics were used to ‘prove’ that a claim that lots of people are having to work long hours and/or multiple jobs to make ends meet was false. The BLS says with a straight face that only a tiny fraction of people work two jobs, and that the average hours per week worked is 34.5. But if you delve into the BLS’s reports, you see that when they collect data on people working two jobs, they only collect data on people working two ‘formal payroll’ jobs. Anyone who works a regular job and drives an uber on the side, or who works a regular job and works in a call center that classes everyone as 1099s, or who operates their own business on the side counts as having only a single job for them. Similarly, survey generating the number of hours worked per week excludes these people, and also treats anyone who is salaried as working only 40 hours per week! So the McDonalds manager who gets ‘promoted’ so they don’t have to pay him overtime and routinely covers shifts for missing employees never counts for more than 40, nor does the IT worker in a ‘burnout culutre’ job who’s expected to work at least 60 hours per week.
While I’m sure the number they report accurately reflects the numbers they collected, what they count for collection doesn’t match what they call the numbers they report. I would be interested in actual numbers on how many Americans work multiple jobs and how many hours per week Americans typically work, but the BLS numbers that claim to be that simply aren’t.
That depends on which ones, and on your definition of “accurate”. Calorimetric values have errors in the 5-10% range; calculations based on them or trying to obtain them are considered perfectly fine if they are within those ranges.
OTOH, we routinely expect any professional scale weighing things which aren’t alive and moving to produce values that are exact to the 4th or 6th figure, or even beyond (depending on the scale, the material and the environment).
And something like pH would fall in between: it’s not as horribly inaccurate as calorimetric values, but anybody giving pH with two decimals should be conscious that the second one is wobbly, and anybody giving it with more needs to go back to school.
the criticisms of these numbers are usually more politically driven also by the ignorance of the practical limitations of the continuous statistical data collection and the types of data collection standardizations that are needed to ensure the continuity of the time series data and the comparability - a very different operation from the single point in time studies or the single point of fixed physical reality data generation used in a physical science like the chemistry etc.
your following examples suggest in fact it is the political point of view generating the complaint.
this is an observation on a usage of a data and not the data.
Here an example - taking as assumption it is an accurate rendition of the American statistic, that reflects the trade offs of the budgets and the reality of obtaining the regular and the consistent survey data responses from the representative samples.
Simplifications and standardizations based on simplifications are necessary frequently to ensure reasonable and consistent response rates within the reasonable limit of the budgets available.
It is also the case the statistical collection (although maybe not the USA with its habit of self centrism) attempts to be standardized with international standards which also have to be made to attempt to ensure the cross comparison
these are questions that are not ones that such data is made to answer.
It is like criticising the hammer for not being the screwdriver.
Typically from both the Left and the Right there is always the criticisisms and the unhappiness if the statistics in question do not have the detail and the orietnation that their current obsessions desire, but this is the problem with the tension between the standardization and the long-term value of the comparability with the evolution of both the political and the policy concerns.
And the accuracy of numbers in the natural sciences can change over time. When the Big Bang first was generally accepted, how long ago it happened was given as 15 billion years, give or take 5 billion years. Now it’s 13.799 billion, give or take 21 million years.
Certainly Politicians do make stats up. Just a brief perusal of Politifact will confirm that.
And I suppose some harried bureaucrat has made up a number he didnt have on hand when pressed.
But you see- bureaucrats have little reason to lie. Their jobs are protected. And of course once they leave, they can simply post tell all blogs or run to a newspaper.
If the conspiracy is limited to a handful of people, maybe. But once it spreads- *someone * is gonna spill the beans, let the cat out of the bag and so forth.