Commissars screwing with the U.S. government don’t want us to see the data

Statistical Modeling, Causal Inference, and Social Science 2025-02-23

On 1 Feb 2025, Dale Lehman wrote:

If you are inclined, I think this is worth blogging about sooner rather than later: A list of government web pages that went dark Friday | AP News

And, if you do, please do use my name! Many of the data sources are ones I have used and contain potentially important information. Regardless of the Trump view of what is good or bad, I don’t think it is justified to remove data because it mentions topics he finds unacceptable. The data is not biased nor does it promote any specific ideology. Of course, it can get used that way, but that is different than the data itself promoting an ideology. And, I have little sympathy for the bureaucrats at these agencies that have yielded to pressure from the administration – I appreciated the dangers for them, but there comes a time to stand up against injustices. If we can’t track things like sexual identify and suicidal thoughts, such issues do not disappear. Whether an analysis is viewed as promoting an ideology should be separate from whether the data is available or not. Lack of distinction between data availability and the use of that data is a dangerous precedent that should upset anybody concerned with data analysis. Of course, any action is but a weak and ineffective protest, but my sense of helplessness is endangering my health. Can’t we at least take a stand?

I agree. For nearly 40 years I’ve been annoyed at government employees not sharing their–our!–data, and it’s even more disturbing when the reason is ideological.

After reading Lehman’s email, I clicked on the link and the missing websites seemed to be online again, so whatever glitch it was, I guess was fixed. Or maybe someone fixed things in response to the news story?

I noted to Lehman, who responded:

Interesting that they are back. Of course, there is now this: USDA ordered to scrub climate change from websites

A bit different – not an executive order, not definitive, but plenty worrying.

He added:

I just checked the Census bureau site. It is back up, but when I tried to access some data on race and ethnicity, I got this:

As with so much of this, it is a moving target. The USDA story I just sent to you claims that a spreadsheet listing all references to climate change (and details) is required for all USDA document/websites. That should certainly improve government efficiency. I’m sure many of the sites and much of the data will be restored – but it is also likely to be sanitized according to the Trump vision of the world. It is hard to blog about a moving target, but I am not reassured.

Then on 3 Feb:

Update: Here is the now present link to the YRBSS site on cdc.gov:

And here is where it currently leads:

The CDC website does say that it is being revised to comply with the Executive Order. But by what stretch of the imagination is historically collected survey data in potential violation of an order to remove all materials that promote a particular ideology? What bureaucrat feels it is necessary to check such things?

I was curious so I checked the YRBSS today (22 Feb) and found this:

This is absolutely nuts. Real Soviet-era crap. Trofim Lysenko would be proud. Because of that court order, I guess, the Youth Risk Behavior Survey Data Summary & Trends Report is still available online. Here’s the table of contents:

Seems reasonable to me. But not to the commissars. I guess they’re bothered by transgender people . . . I searched the document for *trans* and here’s what came up:

During 2023, the national YRBS included a question asking students if they identify as transgender. The national YRBS also included a question asking students how they describe their sexual identity. In this report, sexual and gender identity are represented in two categories:

1. Cisgender and heterosexual. This group includes students who answered that they (1) are not transgender and (2) describe themselves as heterosexual (straight).

2. LGBTQ+. This group includes students who answered that they (1) are transgender or are not sure if they are transgender or (2) describe themselves as lesbian, gay, bisexual, questioning, or some other way.

As seen with other health behaviors, students who identify as lesbian, gay, bisexual, transgender, questioning, or another non-heterosexual identity (LGBTQ+) experienced disparities in healthy dietary, physical activity, and sleep behaviors compared to their cisgender and heterosexual peers.

There was also something on trans-fat and saturated fat, and a citation of a paper, “Transforming Evidence Into Action: A Commentary on School-Based Physical Activity and Nutrition Intervention Research.”

To say, “This page does not reflect biological reality and therefore the Administration and this Department rejects it” . . . that’s just ridiculous. Health surveys ask about lots of questions. They’ll ask about attitudes, beliefs, religious affiliation, all sorts of things that are not biological.

I agree with Lehman when he writes, “Whether an analysis is viewed as promoting an ideology should be separate from whether the data is available or not. Lack of distinction between data availability and the use of that data is a dangerous precedent that should upset anybody concerned with data analysis.”

Lehman’s next update came on 4 Feb:

None of the CDC datasets are available today. Page not found and the main page just indicates that the site is being reviewed to make sure it complies with Trump’s executive order. I would maintain that the existing datasets automatically comply with the order – the data itself does not promote any ideology. The way the data is used could, in theory, fail to comply, but not the data itself. Also, somebody had to make the decision to pull access to that data while it is being reviewed – who are these people? Is it their interpretation of the order or were they ordered to make everything unavailable until it is reviewed? The lack of due process, lack of explanation and detail is simply another way of telling us that we are irrelevant and powerless. I feel like an Orc in Lord of the Rings, just waiting to hear what Mordor says.

I’m sure this will all get sorted out eventually, perhaps even soon. But who is involved in deciding what is consistent with the Executive Order? Are they accountable to anybody other than Trump? Where are the 3 branches of government that you say we have? I only see one.

I get what Lehman is saying on the orc thing–I feel that way too. In answer to his last question: we do have three branches of government, but the legislative and judicial branches haven’t been doing much. I guess the judicial branch did move that webpage online, so that’s something, but it doesn’t seem that all the laws are being faithfully executed.

On 22 Feb, Lehman added this:

Once gain the issue concerns openness and transparency. Musk has claimed that DOGE is completely transparent: “When asked what mechanisms were in place to guarantee accountability and transparency as DOGE ransacks the federal government, Musk replied that “we post our actions to the DOGE handle on X, and on the DOGE website” — clearly referring to the official government URL, DOGE.gov.”

That is from a report. It was dated Feb 12 and I don’t believe applies any longer since the DOGE website is quite there. What I see on that website is a detailed listing of all transactions as well as what are called “Savings.” You can click on any particular transaction and see a pdf of the contract. What I can’t find anywhere on that site is an explanation of how the “savings” were determined. I haven’t gone to the X site to see if it is explained there (though I doubt it) as I don’t like or bother with X or its predecessor, Twitter.

I don’t like twitter either! But I go there every day to post the links to the new blog posts. Recently when I go there I get a lot of Alex Jones videos: doesn’t he owe some kids 50 billion dollars or something?

Lehman continues:

Without any explanation of where the savings come from, this is not transparency in my mind. In a statistical analysis, it is akin to posting the data with some analysis columns, but no explanation of how those columns were created. If I have to guess, I think they did something like I was able to do from the US budget data. For example, when I downloaded all of the US AID transactions I was able to search for any that contained the words “diversity, equity, or inclusion.” Given what Trump and Musk have said, I suppose they consider these “waste” and perhaps eliminating such transactions can then be called “savings.” There are many issues here. Calling DEI “waste, “inefficiency,” and/or “fraud” is a serious misuse of these terms. Fraud is a legal matter and requires much more than an ethical or political disagreement. Waste and inefficiency suggest somebody did a bad job – while it may be true, it requires more than a disagreement with the last administration’s policies, at least in my mind.

But the more serious matter is that DOGE has not provided any explanation of where these savings come from – what exactly is being saved and why? Claiming complete transparency renders the term meaningless – it is just a list of numbers without the explanation behind them. Let me know if you see an explanation somewhere, but my feeling is that it is hidden, not transparent. Similarly, the listings of Real Estate show “saved” for specific properties of specific sizes. I suppose DOGE has determined that these facilities are not needed – but again what criteria were they using. I can “save” a lot of money if I don’t need to explain what I am doing. And, am I completely transparent if I simply provide a list of what I did?

He continues:

Here is an addendum as I think I figured out to read the forms on the DOGE website. For example, here is one savings of $103,398 on a contract of $127,398. https://www.fpds.gov/common/jsp/LaunchWebPage.jsp?command=execute&requestid=239910218&version=1.5 The link is the pdf form linked to the DOGE website. I had thought these were the original contract forms, but they are apparently DOGE forms. This particular one shows that on Feb 10, 2025 it was entered. There is a field for “Reason for Modification” and it says “Terminate for Convenience (Complete or Partial) here: https://www.fpds.gov/common/jsp/LaunchWebPage.jsp?command=execute&requestid=239910952&version=1.5. I haven’t nearly checked all, but that appears to be a very common “reason.” And the DOGE website showing all these transactions is titled “Wall of Receipts: A transparent account of DOGE’s findings and actions.” Here is what the website says:

To me, this is a continuation of the denigration of the idea of reasons or explanations – the antithesis of scientific reasoning. As much as I disagree with many of the actions of the current administration (though far from all), it is the methodology that disturbs me the most. I can tolerate the fact that Trump one the election and is going to do many things I dislike (some intensely). But when facts and reasons are no longer necessary, then there is nothing left for me, except golf.

Unfortunately the two links just above no longer seem to work. I’m thinking that the government should have an official policy to archive all links forever, just as there is a policy (or used to be a policy?) that all government communications be done through official channels such as .gov emails.

I’ve not looked at the government websites in detail; I’m just forwarding the items that Lehman sent me. With all the scary things going on in the world (most of which cannot be reasonably blamed on Trump–or on the Democrats either, for that matter), missing or garbled government statistics have to be one of the least of our concerns. What’s frustrating here is that this is the government not doing its job, for what seem to be ideological reasons or some sort of political gamesmanship.

The Lysenko analogy seems apt to me. I don’t know Lysenko’s motivations, whether they were pure careerism or whether he had some communist ideology mixed in, but he successfully used political ideology as a tool in bureaucratic infighting. I don’t think that anyone in the current government is playing the role of Lysenko, or Stalin; rather, it seems that political ideology is being used as a tool for people to get what they want, and a lack of enforcement of rules is allowing our government to be used as a playground for these manipulations.

None of this is new–the government hired Brian Wansink and Cass Sunstein, after all, indeed I think Wansink served under both Democratic and Republican administrations–but now it seems to be happening all over.