How your company is doing People Analytics poorly (even if you think you aren't doing People Analytics)

You may not think your company is yet doing People Analytics but here is one way your employer is already doing so, poorly.

Employers routinely use Credit Scores to screen candidates.  

 “47% of employers conduct credit checks to screen potential new potential hires” (Society of Human Resource Management) 

That is to say, they use a score designed to measure credit default risk and apply it to employment screening on some dubious premise that this may be predictive of success, good judgment or good moral character in the context of employment.

What is a credit score?

A credit score is a numerical expression of an analysis of a person's credit files, to represent the creditworthiness of the person. Lenders, such as banks and credit card companies, use credit scores to evaluate the potential risk posed by lending money to consumers and to mitigate losses due to bad debt. Lenders use credit scores to determine who qualifies for a loan, at what interest rate, and what credit limits. A credit score is based on a history of financial transactions sourced from credit bureaus. A credit score applies a mathematical "algorithm" to a profile and history of transactions to categorize a loan candidate so that financial institutions can make a better decision about whether or not to loan someone money, and/or to make a better decision about how to price a loan to loan money profitably, with regard to risk.A bank may offer to lend a low credit score candidate a loan but at a higher interest rate. This allows the bank to maintain a certain level of profitability from different risk segments or deny working with certain risk segments all together.

Why do employers use credit scores to screen employees?  

In employment the credit score is presumably used for a similar high level purpose as banks (control risk). In the case of employment you are trying to reduce the risk of a "bad" hire or stated inversely, reduce the risk of not selecting a good hire. Employers have more applicants than available positions and they limited time to consider each candidate, so they want some quick means to make a decision. Or sometimes we make a decision but apply additional rigor through screening devices to prove to ourselves we made the right decision.

Directly from a credit reporting firm’s website : Credit Scores can help employers “make decisions quickly and easily when deciding on potential candidates” (TransUnion)

 To shed additional light on this practice, I bring up a conversation at a recent U.S. Senate hearing…

 “What is the evidence that there is strong correlation between accessing an applicant's credit history and eventually problems of loss to the employer?” Senator Rosenbaum (Oregon)

“We don’t have any research to show any statistical correlation between what is in somebody’s credit report and their job performance or their likelihood to commit fraud.” Erick Rosenberg (TransUnion)

Here are some other facts to consider about using Credit Scores for employment screening:

- 1 in 4 credit reports have been found to have an error

- 1 in 20 have been found to have serious errors

- Other issues : “52% of all debt on credit reports stems from medical expenses” CFPB

 25% error rate? If you thought your core HR data has problems, maybe it doesn't sound so bad now? 

Scores are being used as a reflection of "character" but if the medical debt statistic is correct, in half or more cases credit scores may be low because of uncontrollable circumstances - how is that reflective of character? Imagine that as a result of chance circumstances you are in dire need of money and you are also as a result refused a means of obtaining money?  Does this practice make any sense?  It's predictive but you don't need any math at all to predict this outcome - it's a self fulfilling prophecy - self fulfilling prophecies are convenient if you sell predictions.

How do the credit reporting agencies and employers defending this practice?

The "all else equal" claim. That is that "all else equal" this is a better way of making a decision than nothing.  Is it really? If we can find evidence of errors, systematic bias and NO evidence that there is any relationship with job performance - is this really a better way of making a decision than a coin flip? Or why not do the work to find something else not equal that has less errors, less systematic bias or more evidence of relationship to job performance? Why not find a better way of making things not equal?

We have always done it this way or this is how others are doing it. Come on, is this really a good reason? Is this how to make good business decisions? This is not how any great business decisions are made, ever.

It seems like a "plausible method" of making decisions by logic or rational argument.  

Here are some theories that are at least as plausible as the theory that credit reports may be good predictors of performance …

  1. Maybe because we really have no idea how to screen a candidate for characteristics that relate to performance without doing this work directly …
  2. Maybe because we want to find a way to systematically discriminate against populations that come from impoverished communities, thus discriminating against a high percentage of minorities, without doing so directly… Or we don't delve into the details to see how this math may or may not play out and we don't care.
  3. Maybe because we want to eliminate candidates who have high medical expenses and would cost our health plans money, without doing so directly… (This actually seems like the most mathematically plausible scenario to me)

I ask again, why are you using credit scores to screen job applicants? 

How about this, why don't we actually do the work to see what factors drive performance in our organization and/or isolate with data how we can increase the probability of high performance, regardless of starting individual characteristics.  To me this is a better way of making decisions and a better way to run a business.

Thank you to John Oliver (yes, John Oliver the comedian) for highlighting these issues in a clear, emotionally charged and entertaining way on his show : Last Week Tonight (HBO) – Be advised - wear headphones - there is some language in this that may be offensive and not safe for a work environment. I would argue the practice of using credit reports as screening devices is equally offensive and unsafe.

I wish that each employer would put as much time researching the practice of using credit reports as employment screening devices as an HBO comedian did.

Seems like it would be even more useful to us to know these things than it is for him, is it not?