Ask Slashdot: Can US Citizens Trust Government Data? (msn.com)

mmell writes: An editorial in the Washington Post and made publicly available via an MSN news feed has asked the question: "In the Trump administration era of 'alternative facts,' what happens to government data?" Given that Slashdot members (and readers) may represent a somewhat more in-the-know crowd on matters concerning data integrity and trustworthiness, I thought this would be a good place to ask: can we trust (or has anyone ever really trusted) government data? One might think government data would all be cut 'n' dried and not subject to manipulation, but I personally remember when government data back early in the Reagan presidency went from reporting nearly 15% unemployment nationwide to well under 6% by redefining what "unemployed" meant. So . . . has government data ever been trustworthy, and is it still so?

Comments

Gov't data

[Anonymous Coward]

Have you ever been able to trust it? I doubt it, so nothing has really changed in this regard and the timing of this question seems partisan.

Re:Gov't data

[cayenne8]

More likely the new administration would just stop collecting inconvenient data, or change the metric as described in the summary.

If they did, then they would be merely continuing the SOP for all federal government data collected and reported, as done previously by Obama, and Bush and.....

Nothing new to see here folks.

Re:Gov't data

[DarkOx]

Exactly what so many people seem to missing about all the hubub around presidency is the deep state is real, our bureaucracy for good and ill are quite resilient.

Just because you change out the man at the top and couple handfuls of his direct reports does not suddenly mean all the procedures, methods, systems, opinions, etc in use by all the 2,804,000+ federal workers and enumerable contractors both direct and corporate suddenly change too. That stuff is cultural and other than a few hot button issues that might get attention from POTUS directly takes decades to change, literally outlasting a single Presidents term of office in many cases.

In a lot a ways we are still feeling the effects of not exactly policy but popular opinion that dates to the Clinton Presidency and that of Gorge W Bush. People choose to get into civil service or not often depending on their admiration or lack their of for the top man in charge at the time they are ready to start a career. The people who started their careers in the late 90s and early 2000s are now the folks who have risen to positions where they are decision makers and mid-level bureaucrats. We have yet to see the real influence of Obama's millennial voters here yet (sadly IMHO, not looking forward to that all).

So the data is probably as trustworthy as it was 4 weeks or 4 years ago. Its probably as trustworthy as it was 8 years ago, or 16 or 20. That is to say its really not very trustworthy at all but probably less bias than you might imagine. There is a constant battle being fought between the left and right with the pendulum swinging both ways ever 8 years or so, but not as a far either way as the top men appear to swing. The real issue is that assumptions on either side are never really challenged or well examined because of the tug of war fought over the superficial stuff. So some labor statistic remains calculated they way it has been for the last 40 years when some probably well meaning person made a judgement call based on the information they had at hand. It never gets revisited in a serious scientific way because everyone is to busy doing studies and bickering over a handful of top line numbers that make for good headlines like the employment rate.

Re:Gov't data

[whoever57]

To Trump, facts are whatever he wants them to be.

He is a pathological liar and simply has no concept of truth that normal people would understand. I suspect that when he is lying, he isn't aware that what he is saying isn't true.

Re:Gov't data

[rsclient]

Example of government data I trust:

• Tide tables
• Atomic clock time
• Astronomical and orbital data
• How many people live in different states
• Tabulated results from asking companies how many people they are hiring

Each of these are more or less accurate and precise, and each has their own source of biases, and therefore you can trust them more or less.

Re:Gov't data

[dywolf]

Either way, it sure makes the 5% look suspect.

not after you subtract retirees, student, and stay at home parents.
if you're not working and not seeking work, you aren't part of the labor market despite being "unemployed", and hence not counted.

Re:Gov't data

[meta-monkey]

That's not how pathological liars work. They absolutely know they're lying. But they assume everyone else is lying, too, and it's okay to lie to a liar. Gotta fool them before they fool you, right?

I don't think that describes Trump, though. What he mainly does is exaggerate, or a variation of Cunningham's Law. He says things that are imprecise, but in the direction of truth that otherwise the media would never report. Example, Trump wants people to know "employment numbers are bad." Obama/Hillary says 5%. Trump throws out any random high number (42%, whatever) and then everyone "proves" Trump wrong by saying "well yeah the 5% number is bullshit so maybe it's really like 15% or 20% when you start factoring in underemployed..." etc etc. Okay, great, Trump had a number wrong, but at least he recognizes the problem. Most importantly, the voters who are struggling to find work are now very aware that Trump is aware of their employment problems. They don't really give a shit about the exact number, because the only number they care about is the number of jobs they have. Obama and Hillary also both have wrong numbers, and are wrong about the existence of the problem. If Trump gave the 15% or 20% number himself everyone would just ignore it.

Same thing with Mexican rapists. If he just said "some illegal immigrants are criminals" he would be ignored. So he said Mexico is sending rapists, and then the media has to start arguing over just how many Mexicans are rapists. Fun(?) fact, 80% of central American women and girls are raped during their illegal border crossing. So, as Trump said, "well, somebody's doing the raping." Voters are then more concerned with stopping the rapes rather than nitpicking over the number of rapists.

You just now started worrying?

[CajunArson]

Only a koolaid drinking disingenuous douche-shill thought that the government was magically trustworthy with Obama but all of the sudden is magically not to be trusted anymore because there's a new president.

Especially because it's pseudo-intellectual bullshit since the cancerous unelected unaccountably bureaucracy that actually runs the government doesn't care about who is in charge.

The government is 100% as trustworthy today as it was the during your god and personal saviour Obama's reign. It is left as an exercise to those of us with more than two brain cells to determine what that trustworthiness level is.

Re:You just now started worrying?

[Attila Dimedici]

You make a good point. However, the inability to trust government data goes back to long before Obama. It did get measurably worse under Obama, but that was only possible because the government had gotten away with being progressively more dishonest with its data. At one time, the government intended to be honest with its data. Unfortunately, to some degree, everyone is somewhat dishonest with their data (leaving out data points which appear to be aberrations because they do not fit the gatherer of data's understanding of what is going on, as one example). However, as the government collected more data by more agencies. people discovered they could shape people's actions by which data they released. From there it was a short step to only releasing the data which caused people to act the way those in power wished them to act. Then, there came "normalizing" the data, and making "seasonal" adjustments to the data. Until finally, they just outright changed it to make it fit what they wanted people to believe. What those doing this failed to realize was that at each step on the way more and more people began to distrust the data, until now, people recognize that data released by the government has been carefully chosen to promote the agenda of those who control the government, even if that means making numbers up.

Re:You just now started worrying?

[El Cubano]

Only a koolaid drinking disingenuous douche-shill thought that the government was magically trustworthy with Obama but all of the sudden is magically not to be trusted anymore because there's a new president.

Well put. When I saw this supposed question from the article, "In the Trump administration era of 'alternative facts,' what happens to government data?" I had a similar thought to what you stated. In fact, my first response to the "question" was, oh, do you mean 'alternative facts' like 'Islamic terrorism' not being a real thing, or like that we can pull out of Iraq and be free of our involvement there, or like that we can let Russia come in and take control and that won't have a bad impact on the US or our allies, or how if we just build schools, hospitals and give them jobs, everybody who would have become a terrorist will instead live a happy productive life without perpetrating any violence, or how we should release the bad people from Gitmo because they aren't really bad people they're just misunderstood, or going back to the Clinton administration how we don't have a problem with terrorists that requires a military solution, we have a problem that requires a criminal justice solution?

The list goes on, but you get the idea. All of those 'alternative facts' from Democrat administrations have resulted in the direct and indirect violent deaths of many Americans and other westerners. The Republicans have their fair share, but you can't lay the blame for the problem solely one party.

Re:You just now started worrying?

[Anonymous Coward]

So what? Election's over. Now to eat our dinner.

A president that lies and claims it is very very dangerous and not to be dismissed lightly. This is why a functioning and rigorous press is necessary.

Re:You just now started worrying?

[fastest fascist]

All your complaints seem to be policy points, not facts of any kind ,and certainly not government data.

Re:You just now started worrying?

[squiggleslash]

The list goes on...

...of stuff nobody has ever said?

'Islamic terrorism' not being a real thing - I'm guessing this is a perversion of the Bush and Obama administration's refusal to call IS* or Al Qaeda terrorism "Radical Islamic Terror". But neither regime has said "'Islamic terrorism' not being a real thing", they've said that it's unhelpful and likely to help the terrorists' own campaign if you link the words "Islamic" and "Terrorism" because you're implying that IS*/AQ's linking of the two is legitimate, and because many, including Muslims, will take the phrasing as implying we're at war with Islam rather than at war with terrorists.

or like that we can pull out of Iraq and be free of our involvement there - Nobody has ever claimed this, ever.

or like that we can let Russia come in and take control and that won't have a bad impact on the US or our allies - Where? Syria? Because that's not what I've heard at all. Most politicians on both sides of the fence are deeply troubled by Russia's involvement in Syria. Hence the support for a no-fly zone.

or how if we just build schools, hospitals and give them jobs, everybody who would have become a terrorist will instead live a happy productive life without perpetrating any violence, - nobody has ever made that argument in the history of the universe.

or how we should release the bad people from Gitmo because they aren't really bad people they're just misunderstood - nobody has ever made that argument in the history of the universe. The complaint about Gitmo is two fold: one, there are a lot of innocent people there, and two: it's unconstitutional and illegal to hold people without due process (see (1) for the reason why.) Obama was making plans to move prisoners at Gitmo to the US in the early days of his administration, to US prisons, to be processed by constitutional authorities. After Congress effectively made that option impossible, he did the exact opposite of what you claim: he kept Gitmo open, rather than releasing people being held illegally there.

or going back to the Clinton administration how we don't have a problem with terrorists that requires a military solution, we have a problem that requires a criminal justice solution? - congrats, you found something that's correct, but alas, still not something the Obama administration has ever claimed. Terrorism is indeed best treated as a criminal, rather than military, problem. Turning Terrorists from murderous scum into heroic soldiers is the worst thing the Bush administration ever did, and is probably why we've seen an uptick in terror in general, not just in the creation of IS*, but also in groups not associated with Arabs or Radicalized Religious fanatics such as white supremacists.

The list goes on, but you get the idea. All of those 'alternative facts' from Democrat administrations have resulted in the direct and indirect violent deaths of many Americans and other westerners. The Republicans have their fair share, but you can't lay the blame for the problem solely one party.

None of those "Alternative facts" exists. The nearest you got to, that a twenty year old Democratic government might have had a different view of terrorists to the heroic soldiers view you subscribe to, saved lives - it was a conventional law enforcement operation that prevented the NYE attack on LAX from being better known than 9/11. And it was Bush's refusal to take seriously those policies that lead to 9/11.

And, even if any of those alternative facts did "exist", ie Democrats were actually saying them, none of them would ever cause lives to be lost.

Re:You just now started worrying?

[El Cubano]

All your complaints seem to be policy points, not facts of any kind ,and certainly not government data.

They seem to be policy points, that hasn't stopped the responsible administrations from treating them as facts, particularly when you look at their unwillingness to entertain that their assertions may not be entirely merit worthy.

Of course, if instead I had decided to point out the "Iraq has WMD, we must invade," or "mission accomplished" (regarding Iraq, in 2003!), or "we need the wall street bail out" of the Bush administration, or the "we have to arm the Contras" of the Reagan administration. I probably would have been modded to +5 insightful.

But that wasn't the point. The point was to show that administrations of all stripes will engage in the same sort of behavior when it suits their narrative and political objectives. Heck, Obama himself blasted the Bush administration' method of calculating unemployment as wrongly characterizing unemployment figures as too low, then he didn't change it within his own administration. I'll give you three guesses why that was.

Those were similarly policy issues where the administration deliberately twisted the facts, just like the Clinton and Obama administrations had done with the examples I mentioned in my first post. If you don't believe me, go talk to some folks in the intelligence community who were specifically told what facts they could and could not include in their reports and what the results and conclusions of their analyses had to be. It has happened under every administration at least that I can remember.

Re:You just now started worrying?

[anegg]

Not only should you not trust the "government" data, you shouldn't trust anyone's data, in the sense that you shouldn't accept what people presenting the data claim to be the truth, the whole truth, and nothing but the truth. It is human to game the system; it's built into our minds and into our culture. When we have a message to send, we find the data that makes the best case possible for the message that we want to send, and that is the data that we show.

I'm sure many have heard the saying "... popularised in United States by Mark Twain (among others), who attributed it to the British Prime Minister Benjamin Disraeli: 'There are three kinds of lies: lies, damned lies, and statistics.'" (Wikipedia). At the heart of this is that when you get to pick what data to show, what statistics to publish, you get to control the impression people take away from viewing that data.

If the question is whether you can trust that the actual data presented by a government agency is "good," the answer is probably "yes" as long as you understand all of the assumptions behind the collection and presentation of the data, as well as the particular meaning of the terms being used to describe the data as used by the people who gathered, processed, and are presenting the data. This, in effect, makes it very difficult to take anything anyone with an agenda says at face value, including the government. Two examples of problems with government data come to my mind immediately: Inflation, and Employment.

In the United States, inflation is typically measured by the government using something called "the Consumer Price Index." This represents the change in the cost of a "basket" of consumer goods over time. The goods are supposed to be representative of the goods that most people need to acquire as part of their regular daily existence. https://www.bls.gov/cpi/cpifaq.htm But whether or not this inflation value applies to you and your family depends upon how closely your purchases match the assumptions built into the CPI. To the extent to which you deviate from the CPI, the real effects of inflation upon your purchasing power will vary. Are the people who create the basket of goods used to measure inflation trustworthy? I'm sure that they believe that they are, and I'm sure that they try as best as the can to get it right according to what they think is right. Do they measure it how I would measure it? I don't know, and I would have to understand much about their methodology n order to answer that question.

Employment. Big grey area, as mentioned elsewhere in this thread. Ideally, you would measure the total number of people working, and divide it by the total number of people who would like to be working (assuming for this method that everyone who is working wants to be working), and that would tell you the "employment rate" (and the unemployment rate would be 100% minus the employment rate). Even assuming that you can get accurate numbers regarding how many people are working, its almost impossible to know how many people want to work but can't find a job. Some measures use the number of people who are collecting unemployment compensation, but not everyone who wants a job collects unemployment. Other measures might try to estimate the total labor force using age, etc. Obviously how you estimate this number affects the employment/unemployment rate calculation, quite significantly.

No matter whose data you are viewing, you need to understand the assumptions and the methodology behind collecting, processing, and presenting the data in order to know what that data might saying, and even then you aren't seeing all of the data that isn't being presented (that might give you a more complete picture). Even science has trouble with this, so don't expect anything better from politicians.

Depends on the Department

[IMightB]

I think it depends on the dept that your talking about. I wouldn't trust anything coming from Trumps office for shit... Places like NOAA, NASA, etc etc. I would probably trust more. It all depends how how horrible it gets under Trump.

Why restrict this to US citizens?

[OzPeter]

Trusting government data is an age old problem, and even though I might Godwin myself over this, Goebbel said things like:

A lie told once remains a lie but a lie told a thousand times becomes the truth

and

The truth is the greatest enemy of the State.

That Trump is trying to channel Goebbels* is not surprising given who Trump is and his past utterances. (But I don't know which Emperor he was trying to channel when he proclaimed the "National Day of Patriotic Devotion" which coincided with his inauguration - seriously .. it's a real thing)

* The headline of TFA is "In the Trump administration era of ‘alternative facts,’ what happens to government data?", something that TFS should have taken into account.

The questioner reveals their own dishonesty

[DeplorableCodeMonkey]

One might think government data would all be cut 'n' dried and not subject to manipulation, but I personally remember when government data back early in the Reagan presidency went from reporting nearly 15% unemployment nationwide to well under 6% by redefining what "unemployed" meant. So . . . has government data ever been trustworthy, and is it still so?

Under Obama, we stopped counting people as unemployed if they gave up looking for a job. Such people are difficult to track is the argument. Oh really, that's a rational argument for moving the goal post? Controlling for some variables is really hard, so let's not? Well fuck me, I guess we're all Barbie now and "Math is hard!!!"

The inflation measurements have long been known to exclude food and fuel, the two most inflation-affected consumer goods. The food pyramid was laughably unscientific when it was created. Need I go on?

You got Trump in no small part because of this faux earnest, "we're just interested in the facts, sir." No, you're not. You're as fucking dishonest as Ellen DeGeneres when she said "8 years, no scandal." Please ignore Fast and Furious, the fact that Clinton intentionally lied about the known motivations for the attack on Benghazi, political appointees at the IRS targeting conservative groups, blowing up a 16 year old kid with drone and so much more.

Silly question

[silentcoder]

Government data is a collective noun for reports from hundreds of different organisations, some run by people obsessed with clerical accuracy and some run by people obsessed with partisan propaganda who are happy to massage a number. All of them under two houses of congress and a white house which changes its stance on what numbers they wish were different and how on a biannual basis.

The result is that you simply cannot ANSWER the question "is government data reliable" - there's just no single answer. A lot of it is reliable and, in fact, the best data available on some topics. A lot of it is flagrant bullshit, or at least deliberately presented in a way to deceive. And then there is data where it's more ambiguous. A lot of Austrian economists disagree with the inflation rate - and claim it's much higher by doing a different calculation from the raw data to substantiate this - while a Keynesian mainstream economists generally agree that the official figure is a decent representation of the number. It may not always accurately reflect price shifts (and it's always a bit behind the times) but for economic policy decisions it gives the information that is needed to make decisions like "should we raise interest rates, are we in a liquidity trap that demands quantitative easing or are we in a boom-cycle where that will cause a disaster ?"
Who is right, will largely depend on whether you think Austrian economics is a cult divorced from any usefulness by it's refusal to accept empirical data as evidence and thus happy willingness to reject the constant failures of it's policies to have the right results as evidence against those policies... or see Keynesianism as a rampant scam designed to give government the power to decide what money is worth and control everybody's lives (I subscribe to the "The version of economics best supported by empirical data and historic ability of it's predictions to have expected outcomes is the most scientific" school - which is Keynesian through and through).

Some government data is the result of strenuous scientific study which is highly unlikely to be false, fabricated or manipulated (and almost impossible to apply to do this with), a lot are from softer human sciences which is more susceptible to this.

There is no answer to the question of "is government data reliable" - but you CAN answer "is *this piece* of government data reliable".

It's interesting how the Donald seems dead-set to pursue his agenda not by altering government data (particularly the scientific type) but by eradicating it - defunding or abolishing government research agencies that produce data on topics he would rather pretend is different or non-existent. As is climate change will stop happening if he defunds NASA's earth-science division so they can't tell us about it anymore. Sure this will weaken science over-all by removing a valuable source of data on how fast things are happening, but it won't make them stop happening. That gives you a clear view on the difference between easy-propaganda-data and scientific-data. Trump is well aware that he cannot pressure NASA to start reporting denier-friendly results, they are too well scrutinized by other scientists outside the agency, and if they suddenly stopped publishing raw data it would look too suspicious - so his best answer to keep his claims from being challenged by his own agency is to silence the agency.

Re:We could never trust government

[AmiMoJo]

In a properly functioning government, independent bodies are created to gather data for use by the public and politicians. Those bodies are overseen by bi-partisan groups with representatives from multiple parties, and their mandate is independence, transparency and impartiality.

It works in some democracies. The UK has the Office for Budget Responsibility, Japan has various agencies... Of course, politicians do their usual thing of cherry-picking and and misinterpreting the data, but the raw stuff is available and generally considered reliable.

It's also not the case that US governments have always been this bad either. I don't recall one telling the National Parks Service to stop publishing factual information because it contradicted their lies over something as trivial as the size of the inauguration crowd.

Before you think about this, answer me that:

[Opportunist]

What changed during the Bush administration?
What changed during the Obama administration?

Why the fuck do you expect change now?

I don't get it. I really don't. NOTHING changed from one administration change to the next for the past decades. Oh yes, there was a war on terror. Oh. And? That would have been in what way different under any other rule?

Face it, folks: You're fucked. You have a system in place that allows you to choose every 4 years whether you want to feed one group of useless gits or the other group of useless gits. That makes a huge difference for the gits, and that's why that election fight is fought tooth and nail because it's all or nothing for them. Fo you, it's nothing. Either way.

Mostly because you don't get to choose who you can vote for. That's chosen for you. In the end, when you strip the whole fluff, the whole spectacle has a lot of the old Soviet times when you even sometimes got to choose between two candidates from the same party, supporting the same ideals and the same economic system, not questioning in the slightest the all-holy doctrines and differing in insignificant bullshit topics that were hyped and emotionalized to insane levels despite having exactly zero impact on anything that really mattered in the end.

Let's be brutally honest: The same is true for your DemRep Party.

Re:We could never trust government

[AmiMoJo]

The two party system is a problem, but it's nothing compared to the danger of what Trump wants.

Trump is pushing for people to accept his "alternative facts" over the truth. It's a standard feature of post-truth politics. You pick the "facts" that you prefer. He wants voters to do that, to believe him rather than the press or government agencies that are publishing contradictory information.

Both Trump and his press secretary and various members of his staff have said this over and over. The press is dishonest, anyone who contradicts him is a liar. It's extremely dangerous to disregard the truth and stop caring about it.

Re:We could never trust government

[whoever57]

Is that another Trump fact?

McCarthy was not "completely and totally vindicated" by anything. The existence of a small number of Soviet spies did not justify the witchhunt that he pursued for political reasons. It did not justify all the false allegations and innuendo, it did not justify ruining of the careers of many innocent people.

Re:Who Has EVER Trusted Government Data?

[hey!]

Who trust government data? Anybody who uses a USGS map. Or a weather forecast that uses satellite data. Or who uses a GPS (both the satellite signal and the base map, which is compiled by private companies from government sources).

Now any statistic is capable of misleading, if you choose to misinterpret it. Take unemployment. I think that figure is accurate, it just doesn't mean what people think it does. By 2016 unemployment had recovered to where it was before the Great Recession, but if you think that means the government is fraudulently telling you that the job picture is good, that's you misinterpreting what it means. The low unemployment rate masks (a) relatively low labor participation and (b) disastrously low job growth and labor participation in certain regions of the country -- particularly rural and small to middle-sized cities. How do I know this? Well, government data, obviously.

You are conflating "data", with "information" and "opinion". The Food Pyramid is opinion, not data. If you think for yourself and drill down into the facts a bit, you'll find that government data is pretty useful. Opinions, less so.