The Obama Department of Labor in its tenth year

I’ve written a series of posts about how the Department of Labor under Alex Acosta has barely lifted a finger to overturn the radical policies and practices of the Obama DOL. Acosta has left former president Obama’s Administrative Review Board, the DOL’s appellate court, in place. He has refused to disturb the pro-illegal immigrant polices of Obama and former DOL Secretary Tom Perez.

Acosta changed the Obama administration’s interpretation of independent contractors under the FLSA with respect to home health registries. However, to the consternation of Sen. Marco Rubio, he permits career employees to continuing using the Obama administration’s interpretation.

No wonder those who deal with the Acosta Department of Labor refer to 2017 as Year Nine of the Obama DOL.

Today, I want to discuss what may well be the most egregious respect in which the Obama DOL persists. It involves the Office of Federal Contract Compliance Programs (OFCCP).

The OFCCP monitors federal contractors to ensure that employers doing business with the federal government comply with laws and regulations that prohibit discrimination. Whether the OFCCP finds federal contractors in compliance with anti-discrimination laws depends, of course, on how it views discrimination.

In our system, Congress, not the federal bureaucracy, is charged with determining what constitutes unlawful discrimination. In practice, we rely on courts, not federal bureaucrats, to flesh this out.

However, because of the OFCCP’s power over federal contractors, federal bureaucrats can apply their own views of anti-discrimination law to a vast swath of the nation’s employers, imposing penalties (and potentially even the loss of the right to do business with the feds) on those deemed out of compliance.

Contractors can litigate against the OFCCP, but they do so, in the first instance, before the DOL. Their right to appeal lies, in the first instance, with the Administrative Review Board where the decision will be made by the committed leftist Alex Acosta has kept in place.

There is a right, eventually, to litigate in federal court. But considering the cost of that much litigation, and its uncertainty, many contractors take the path of least resistance and bow to the OFCCP. Those that don’t pay a big price even if they ultimately prevail.

Fortunately, the OFCCP’s interpretation of the law has not, traditionally, been wildly out of step with mainstream notions. Unfortunately, but not surprisingly, that changed quickly and dramatically during the Obama administration. Surprisingly, it has not changed back in the Trump administration.

The clearest manifestation of this problem is in the area of compensation. Again, this is not surprising given the left’s obsessions with the ( mostly mythical) pay gap — the debunked claim that women receive 77 cents for every dollar a man earns, and all that.

The Obama administration decided that OFCCP would be the vehicle through to address the “pay gap.” From its perspective, this was a smart decision. Using the OFCCP would eliminate the need for complaints by actual females (too benighted or oppressed to know they are being discriminated against) and remove federal courts from the equation, at least in the first instance. With federal courts sidelined, Team Obama could push radical theories of compensation discrimination with the reasonable hope of making them stick.

Obama’s new Secretary of Labor, Hilda Solis a congresswoman from California, appointed Patricia Shiu to head the OFCCP. Shiu was an plaintiffs’ employment lawyer from the San Francisco Bay area. She staffed up for her crusade by placing two plaintiffs’ class action lawyers in key positions (and giving them civil service jobs).

After informing the contractor community that “there’s a new sheriff in town,” Shiu and her posse immediately set out to find pay discrimination as defined by their radical conception of it. Their weapon was statistics.

Statistical evidence can be a helpful tool in spotting possible discrimination, but only if statistics are not misused. Shiu’s OFCCP willfully misused them in ways I’ll describe in a moment.

Before doing so, I want to emphasize that, with Shiu gone and Acosta in charge of the the DOL, OFCCP continues its indefensible use of statistics to find pay discrimination where it doesn’t exist. Just before exiting at the end of the Obama administration, Shiu brought a series of actions predicated on the radical misuse of statistics. Normally a new administration run by a different political party will hit the pause button on its predecessors action, especially on its “midnight” moves. For example, the National Labor Relations Board did this after Trump took office.

But there has been no pause at Alex Acosta’s DOL. The OFFCP is still running with Shiu’s midnight actions and the key people Shiu brought in, and gave civil service jobs to, are still running the show.

Back to statistics. The main abusive practice of the Solis-Perez (and now the Acosta) Labor Department is the aggregation of dissimilar jobs for comparison of pay via statistics.

When comparing male and female pay rates, it’s vital to compare the pay of people who are performing the same kinds of work. For example, in the tech industry, a prime target of the Solis-Perez-Acosta DOL, it makes sense to see whether male and female engineers performing highly complex work (e.g., on the cloud or on artificial intelligence) are paid about the same. If they aren’t, the contractor should have to explain why.

But it makes no sense to lump all people holding the title “engineer” together. One would expect engineers performing sophisticated work to be paid significantly more than those performing relatively unsophisticated work, such as tweaking Outlook. Thus, no inference of pay discrimination arises from pay differences within such a broad classification.

Moreover, it is preposterous to aggregate all “exempt” employees and compare their pay. Of course, the CEO of a tech company will be paid more than engineers, and engineers will be paid more than, say, human resource specialists. Inferring discrimination, or even the hint of it, from gender pay disparities within a group this broad is absurd.

The Solis-Perez-Acosta DOL has aggressively aggregated people holding the same job title in its statistical analyses of pay. OFCCP’s case against Oracle, one of those “midnight” actions, provides a good example. The Chamber of Commerce, in a Fall 2017 report on OFCCP states:

OFCCP bases its allegations of discrimination solely on statistical analyses and offers no other evidence of discrimination. In doing so, OFCCP relies on Oracle job titles to attempt to demonstrate that certain female, African-American, and Hispanic employees were “comparable” to males and Whites employed in similar roles.

But job titles cannot serve as proxies for establishing that certain employees are proper comparators under federal antidiscrimination law. To OFCCP, it appears to be all about the numbers that it can spin from aggregated employer-provided data. Individual employee differences relating to productivity, required skills, and experience are ignored. This issue remains pending as of this writing.

It also remains as of this writing, months later and more than a year into the Trump administration.

Under the Trump administration, the OFCCP is taking its abusive aggregation practice one step further. I’m told that it wants to use an approach that would aggregate employees based on the qualifications, skills, and experience — the “human capital factors” — they possessed when they first applied for work with the contractor. Everything else that might explain pay differences — e.g., what jobs the employees pursued; what additional skills, qualifications, and experience they went on to obtain; whether at some point they chose to work part time — is deemed irrelevant for purposes of finding pay discrimination.

The notion is that once two comparable individuals apply for work at a company, everything should be the same thereafter. If it isn’t, the theory seems to imply, the company is to blame.

The OFCCP also wants to discount the market. A federal contractor needs to pay the market rate to engineers working on artificial intelligence; otherwise they will work elsewhere. The going rate will be significantly higher than for an engineer performing mundane work or for an HR coordinator. To the OFCCP, such market variables, and the gender pay differences they produce, are evidence of societal discrimination. Thus, increasingly, it looks askance at reliance on market data to explain gender pay disparities.

As my first boss at the EEOC, a pioneer female government lawyer, told me, there’s a name for these of approaches to compensation issues: “socialism.”

The passage from the Chamber of Commerce I quoted earlier regarding Oracle points to another problem with the OFCCP’s approach to statistics — OFCCP now finds discrimination based solely on statistics. As the Chamber stated:

The Palantir and Oracle matters—like the Google matter— are a few examples of the many cases in which OFCCP is pursuing claims of discrimination based on statistical evidence alone. Indeed, the Agency’s practice has become so common it has not gone unnoticed by Congress. House Report 114-99 accompanying the appropriations bill for the Departments of Labor, Health and Human Services and Education, issued in July 2016, expressed deep concern over OFCCP’s overreliance on statistical “evidence” of discrimination.

This concern is as great as ever under the Acosta DOL.

In cases of extreme statistical disparities courts will infer discrimination from numbers alone. Normally, though, anecdotal evidence is needed to buttress a statistical case.

But the OFCCP will find discrimination based on a showing that a contractor’s numbers are two standard deviations from the expected one — a showing that, assuming the comparisons are apt (and we have seen that they typically are not), means only that there’s a five percent or less probability that the disparities are the result of chance. The OFCCP does not require anecdotal evidence. Its investigators have acknowledged, informally, that such evidence often cannot be found.

OFCCP’s view of the use of statistics to prove discrimination is out-of-line with most of the case law. Tellingly, the Agency has refused even to participate in an EEOC-led task force that is considering this subject.

The other big problem with the OFCCP’s use of statistics is lack of transparency. Not only does the OFCCP refuse to show its statistical findings to contractors it accuses of discriminating, it often refuses even to disclose its methodology.

As the Chamber of Commerce report explained, in compensation cases the OFCCP “for decades. . .identified the incumbents or job titles at issue and provided the contractor with an opportunity to explain differing pay among employees OFCCP viewed as similarly situated.” But now:

OFCCP—as a matter of course and enforcement posture— issues a NOV [notice of violation] out of the blue and demands a response by the contractor within a few weeks. The NOV often lacks details as to the statistical methodologies used and the assumptions made by the Agency.

Contractor requests to discuss the NOV in detail are often denied, with the Agency insisting upon a written response before any explanatory meeting is held. When OFCCP finally agrees to meet regarding the NOV, the information shared remains scant.

All some contractors get from OFCCP is an assertion of pay “inequalities” and a demand for millions and millions of dollars to redress them. When contractors try to conciliate with the OFCCP, they often are told to just pay up. This unwillingness to engage continued throughout 2017, the ninth year of the Obama DOL, and it continues still.

There are other serious problems with the Acosta OFCCP. They include absurdly sweeping demands for data and unlawful collaboration between OFCCP officials and plaintiffs’ lawyers.

The Google case is a perfect example of OFCCP’s ridiculous data demands. The Chamber of Commerce report describes this matter beginning at page 26.

OFCCP made its oppressive demands of Google just a few weeks before Trump took office. The DOL continued to pursue them during the Trump administration.

An administrative law judge rejected the government’s position and criticized the OFCCP for taking it. But the Acosta DOL has appealed the decision.

The appeal will be decided by the Administrative Review Board which consists exclusively of leftist appointees whom Acosta has refused to replace, despite having full power to do so.

Such is the state of play as we move through the early months of Year Ten of the Obama Department of Labor. There is no new sheriff in town — at least not one who is yet willing to rein in the old left-wing hangin’ posse at the OFCCP.