In April, I had the distinct pleasure of presenting at DisruptHR Denver. My topic? Drug testing and marijuana. Take a look and let me know what you think! (I mean it, tell me what you think!)
Yesterday, Reuters reported that Amazon created a recruiting engine using artificial intelligence. This isn’t news. Amazon is a leader in automation, so it makes sense that the retail giant would try automation in their own recruiting processes to try to quickly find the “best” candidates. Yet, Amazon’s tool had a big problem – it didn’t like women.
As the article describes, “Everyone wanted this holy grail,” one of the people said. “They literally wanted it to be an engine where I’m going to give you 100 resumes, it will spit out the top five, and we’ll hire those.” Who doesn’t want this? To make hiring faster and easier? Currently, there are hundreds of AI tools available to human resources – many of them in the recruiting space – that promise to do these things for you. But if Amazon found problems, what about those tools?
Amazon’s tool used a 10-year look back of existing employees (largely male-dominated). The tool then could rank applicants based on what it learned makes a good Amazonian. Based on its own analysis, the tool learned that male candidates were preferred over female candidates in a mixture of words that appear on applications, like “women’s,” experience, job requirements, and potentially proxies for gender. While Amazon tried to solve for this problem – making “women’s” a neutral word so the tool did not reduce the applicant’s rank – the results of the tool still had a negative impact on women. So, in 2015, Amazon abandoned the tool. Good for Amazon. This is the right thing to do. But again, there are hundreds of other AI tools out there.
- We know that AI is biased because people are biased.
- We know the sources of the bias include the data we use to teach the AI, the programming itself, the design of the tool, and people who create the tool.
- Employers have to be vigilant with their tools. We have to test for bias and retest and retest (and retest) for bias in our tools.
- Employers – not the AI – are ultimately responsible for the results of the tool, because even if we follow the output of the tool, the employer is making the ultimate employment decision.
It is very possible, even probable, that the tools out there on the market have bias in them. Employers can’t simply rely on a vendor’s salesperson’s enthusiastic declarations that the tool eliminates bias. Instead, employers should assume bias plays a factor and look at their tool with a critical eye and try to solve for the problem ourselves.
I applaud Amazon for doing the right thing here, including testing its tool, reviewing the results, and abandoning the tool when it became clear that its bias played a part the results. This isn’t easy for every employer. And, not every employer is going to have the resources to do this. This is why employers have to be vigilant and hold their vendors accountable for helping us make sure bias isn’t affecting our decisions even when using an AI tool. Because ultimately, the employer could be liable for the discrimination that the tools aid.
There are two questions that can change how well our people perform, how we work as a team, how we manage, and how we keep compliant. Here they are:
- How are things going?
- What can I do to help you?
Definitely not rocket science, but think about these. If you manager came to you, and genuinely asked, “how are things going?” how would you respond? Would you respond with some of your concerns or roadblocks, would you say “my mom has been really sick” or “I’m having a hard time getting through to my Assistant,” or would you say “I completed this project!” More likely than not, if you believed your manager really wanted to know, you’d share information about your or your team’s work performance. You might also share information that affects that work performance.
If your manager asked what she could do to help you, would you give an honest response? “Janelle in Accounting is holding this up, could you please chat with the CFO?” “I would like to go to this conference so I can learn more about XYZ.” “I might need your help filling in for me while I get my mom to the doctor.” Or, “James has been saying weird things to me, could you help me figure out how to handle the situation?” If you know your manager is willing to help, would you ask for it? Wouldn’t this help you?
The Harvard Business Review published an important article about questions and how they build emotional intelligence and most importantly, trust. If all the research is correct that when employees trust their manager, their performance and engagement increase, why wouldn’t we ask managers to ask questions to build trust? These questions are business related by identifying successes and concerns while offering to help.
So, how does this tie to compliance? Well, that’s an easy connection – when would people trust us, they tell us when something isn’t going quite right. They tell us when someone said something he shouldn’t have, when they need a reasonable accommodation, or when they fear a co-worker might be breaking the law. If we want to foster communication from employees on these issues, we need them to trust us. So, let’s ask them the two questions more often.
One other thing – it’s easy to train managers to lead with these questions. The hard part is getting those managers to live these questions, to turn them into real information-seeking questions. Look for those managers who do it well, keep them, train them, promote them.
Photo by William Stitt on Unsplash
HR has two faces. This is not necessarily a bad thing – it’s by design. To employees, we are the face of management. To management, we are the face of employees. This is what makes our job so darn hard.
Employees need us to be their advocate to management. We beg, borrow, and steal for increased benefits, argue that people should be disciplined, we ask for more training and opportunity for employees, we pass tissues when someone is upset, we help get them leave when a grandma is sick or a baby is on the way, etc.
Management expects us to keep them safe. We create policies that govern employee behavior, we find people so the widgets get made, we draft severance agreements when a manager screws up, we coach managers how to have tough conversations, we try to reduce liability by implementing safety programs, etc.
Some of us really like one side over the other, but both are expected and important. Problems arise when we show the wrong face. When we brag to employees just how hard we’ve been working on their behalf, management hears of it and loses trust in us. When we gossip with employees about management strategy, employees learn they can’t trust us either. When we stop advocating for employees with management, the employees know we’ve stopped. Then, they know they can’t come to us with problems – we’ll just can’t be bothered by them.
These two faces are really hard to maintain. We can’t have deep friendships with the people we work with for fear of the time when we have a tough conversation with our friend that ends in her termination. But we want to know enough about people so they feel comfortable coming to us with concerns even if we don’t partake in the party after the company party. This professional distance is important for us to do our jobs, so we’re trusted enough to do our jobs.
Sometimes, we can’t show either face until we have facts. When we hear about harassment is one of those times. We are empathetic and appreciative to the person who brought us the complaint, but we can’t make admissions like “I’m so sorry” or “We’ll make this right.” This could show the employee that we’re on her side, undermining our investigation. OR – perhaps more destructively – we show the employee our management face that’s cold and “just the fact’s ma’am” ensuring that no employee ever tells us about harassment again because we appeared to immediately take the side of the harasser. Once we have completed the investigation, we advocate to management about what should be done, even if it’s nothing. When we advocate for the ultimate penalty (termination), we put our credibility and relationship with management on the line. Then, if we lose, we have to make tough decisions for ourselves.
We don’t always win. We advocate and lose on both sides, employees and management. But even when we lose, our job is to build relationships strong enough to get past the loss so we can be ready for the next round. OR, if this loss affects our integrity, we have even tougher decisions to make. We must think about the possibility of making that decision someday. It takes guts.
We may have two faces, two jobs that sometimes conflict, but that’s the great thing of HR. We get to do both.
Photo by Soroush Karimi on Unsplash
While I have been safely ensconced in #SHRM18, I haven’t been able to read the news as much as I’d like. When I finally looked at my twitter feed devoted to news, I became angry, sad, frustrated, and a whole other host of emotions. So as midnight approaches, here are some things I hope all of my HR friends take from this fantabulous conference to put into their worlds:
Compassion. Oscar Munoz explained why caring comes immediately after safety at United. Caring means holding a door open for a family who just landed a half a terminal away and who are running to catch the plane to see a sick grandma. While a policy may say one thing, caring about the people we serve (and for those of us in HR, that includes our employees and candidates) sometimes says something different. If our employees are empowered with compassion, they will do the right thing for our customers, clients, and the greater world.
Compassion. While he may not have said it in quite this way, Tim Sackett talked about how CEOs want to be able to personalize our HR plans because our people are individuals who want personalization. Personalization means we have to know, acknowledge, and understand the needs of candidates and employees. We can’t personalize unless we are compassionate with the people we help every day.
Compassion. In discussing inclusion, Joe Gerstandt asked us to imagine a world where employees have space to be themselves, we ask and they speak about the personal parts of their life so they don’t feel they have to hide parts of themselves. “How are you really?” “How is your mom? Is she feeling better?” Adding circle tables to a break room so people can interact. Integrating our values into conversations about our objectives, especially when we are struggling with an issue. We want our employees to be innovative problem-solvers, and we can do that by being compassionate with them.
Compassion. I was unable to attend Adam Grant’s presentation. But from what I saw on the twitters, it was amazing. One thing he challenged me on is ending exit interviews. The argument (via him and some super HR pros) is that we should have known about the problems before the employee leaves. This is absolutely true. We should have known. When an employee is so afraid to talk to us while still working for us, we have lost. Lost big time. We need employees to want to talk with us, to want to share the good stuff and the bad stuff. This takes trust. We can foster trust by being compassionate with our folks. Knowing their names, their struggles, their successes. When they see that we are interested and invested in their well-being, they will come to us with their concerns.
So, what do we owe each other? Do we owe each person around us respect? Hells to the yeah. Do we owe each other attention when a problem crops up or a success is achieved? Yes. Do we owe someone time when he is asking for help in dealing with FMLA paperwork because his wife is ill? Yes. All of this takes compassion. When we see people suffering, do we owe them help? Yes. It breaks my heart to see people suffering. I hope that is true for everyone in HR. We owe ourselves, our employees, and the people around us compassion.
I’m going to try to remain hopeful and do better myself.
Photo by Matheus Ferrero on Unsplash
While I totally loitered at the Society for Industrial and Organizational Psychology Conference (I was a presenter, just failed to register – oops), I’d thought a post on what we talked about yesterday and a bit about what’s happening at the University of Minnesota’s HR Tomorrow Conference today: adverse impact, why it’s important, and why you should care.
Adverse impact (known as “disparate impact” by the lawyers) is when groups of individuals described by a particular characteristic is negatively affected by an employer’s decision, selection tool, or policy when that decision, tool, or policy is neutral on its face or does not intend to actually have a negative impact. For example, if an employer uses a psychological test that filters out African Americans, the test would have an adverse/disparate impact on African Americans.
The concept of disparate impact has been around for a long time. The United States Supreme Court in Griggs v. Duke Power formally recognized the claim. Since that time, the law has been debating many aspects of the claim, including what statistical models to use, does the doctrine apply if the rule intends to discriminate, how does impact different from treatment, and will the doctrine apply to all the HR technology out there. While this post could go on-and-on about all of these questions, this last piece is really important for HR tech buyers, and the answer is probably.
We already know that lots of HR technology vendors, including the fancy-dancy stuff like artificial intelligence, machine learning, algorithms, etc., market their products as the only way to find the best candidates, identify problem employees, and make all your dreams come true. When these technologies are used, their use could create a disparate impact. How do we know? Because we’ve already seen how these technologies discriminate outside the world of HR – see photo ID that classifies African Americans as gorillas, recidivism tools that increase prison terms for African Americans, etc., so it is highly likely that they could operate the same way when it comes to HR tech. Arguably, HR tech has the potential to greatly impact because the decisions HR makes affect individual’s livelihood.
So what should we do about diverse impact? While there are many, many things we need to do to limit the potential that the HR tech we use doesn’t discriminate, we should start with two things. First, we have to know how the technology works and the data it uses to make recommendations. This requires vendors to be open and honest with us, lose the marketing gloss, and really explain their products. Can they explain how the tech works? Can they explain how the tech works on our organization’s data? Could the data have bias baked in? (The answer to this last one is probably yes, especially if we’re looking at hiring or performance data. There’s just no escaping it.) When vendors are transparent and honest about these issues, we can take more steps to mitigate any disparate impact the tech might have.
Second, we need to test and test and test to see if the tech creates the disparate impact. Lawyers and data scientists talk about validation as the test. For lawyers, validation means under the Uniform Guidelines for Employee Selection Procedures. For data scientists, validation means how strong the correlations are statistically. This definitional problem causes more debate and potential confusion. So, we need to find vendors who understand, appreciate, and can articulate validation under both tests. Because the HR tech world is a bit like the wild, wild west, it’s hard to find them. (Trust me, they’re out there. I’ve probably met them or at least brow-beat them from a distance on this very issue.)
All that said, I want HR to understand and appreciate that these issues could exist and start playing an active part in fixing these issues. While I’d love for everyone to trust each other, placing blind faith in a vendor is not in our organizations’ best interest. Holding people accountable is one of the strengths in HR. We should use it here too.
One final note, I love this stuff. This tech is going to revolutionize how we do business. I just want to do it in such a way that doesn’t create that much risk for our businesses. Remember my pledge?