We've been running a data science training program for young people with non-traditional tech backgrounds. This training program has been highly successful in some ways - we have an over 80% completion rate on our highly technical MOOCs that cover Github, Version Control, R, Data Communication and more. The young people who complete the program have shown significant skill despite the compressed nature of our program. Given that it is a workforce training program and not an accredited degree, the folks who complete our program come out with a basic set of technical skills, but still have a lot to learn.
The place where we have found more challenges in the establishment of our program is getting our program completers into entry level data analysis, data entry, and data management jobs. Our completers definitely have the skills to take on these rolls, but they don't have traditional qualifications. This has presented significant barriers for us in finding employment opportunities for the young people we work with. Some of the challenges in creating employment opportunities have been due to the fact that data science is a pretty diffuse discipline with different definitions at each company. Some are due to the relative inexperience in the workforce of our trainees and the challenge of working in a new environment.
These topics deserve their own posts, but a lot of the barriers are due to human resources.
I have probably spent the last three years thinking more about human resources practices than any other biostatistician on the planet with the possible exception of Travis Gerke. In trying to generate employment opportunities for our scholars we have encountered all sorts of challenges that I think are worth discussing, because they represent real barriers to careers for young people in our training program - but also are likely contributing to barriers preventing access to the technical workforce more broadly.
In keeping with the theme of these posts I'm just going to informally list barriers we have run into so they don't stay couped up in my head, but I would love to continue the discussion with anyone who is interested.
The place where we have found more challenges in the establishment of our program is getting our program completers into entry level data analysis, data entry, and data management jobs. Our completers definitely have the skills to take on these rolls, but they don't have traditional qualifications. This has presented significant barriers for us in finding employment opportunities for the young people we work with. Some of the challenges in creating employment opportunities have been due to the fact that data science is a pretty diffuse discipline with different definitions at each company. Some are due to the relative inexperience in the workforce of our trainees and the challenge of working in a new environment.
These topics deserve their own posts, but a lot of the barriers are due to human resources.
I have probably spent the last three years thinking more about human resources practices than any other biostatistician on the planet with the possible exception of Travis Gerke. In trying to generate employment opportunities for our scholars we have encountered all sorts of challenges that I think are worth discussing, because they represent real barriers to careers for young people in our training program - but also are likely contributing to barriers preventing access to the technical workforce more broadly.
In keeping with the theme of these posts I'm just going to informally list barriers we have run into so they don't stay couped up in my head, but I would love to continue the discussion with anyone who is interested.
- HR filters on qualifications, not on skills. This means that no matter how skilled our employees are they can't get past the first stage of a screening process. Some of this is unavoidable because of important legal protections like the Fair Labor and Standards Act, but some is due to lack of flexibility in job classification.
- An example of this is our concrete experience with JHU. We tried to hire our program graduates directly at JHU but there was no job classification available in the HR system that could be applied to data science roles without a bachelor's degree. Despite a supportive HR department, it took us almost a year to get a role defined that was appropriate for technically skilled, but not traditionally qualified roles.
- Entry level positions aren't entry level. This is an issue that I know goes far beyond our program. But entry level positions often expect candidates to have significant experience in the area. So rather than viewing a new hire as a long term investment in the potential of a new candidate, they view it in terms of immediate efficiency. This means that our scholars need significant experience to be able to start entry level jobs.
- HR moves very slowly. Often the folks we work with don't have a job (even before the pandemic) and are in urgent need of opportunities. But we have seen delays from three weeks to multiple months just to get an application reviewed for a job where we know the candidate, have the funds, and want to make the hire directly. A variety of obscure bureaucratic steps and overtaxed staff slow down even the most direct hires.
- HR practices are obscure. An example is the above mentioned job category. There is no easy way for hiring managers to access all of the job codes in an institution as big as JHU. This means that it is not possible to find the most job codes without lots of back and forth with already busy HR professionals.
- Technical jobs lack classification flexibility. In the US jobs are classified into two main types - exempt and non-exempt. Exempt means you are exempt from paying overtime and minimum wage - but these roles are typically salaried and higher paid. The FLSA puts in specific requirements for these exemptions to prevent abuse by employers (this is good!). But it also means that if you are doing a job like data science and you don't have a bachelor's degree, you don't meet any of the exemption qualifications. Why is this a problem? If you have technical skills but aren't a computer programmer, all of the technical jobs will be listed as exempt - which is a classification you do not qualify for.
- HR has policies that are unevenly applied. Some examples might include certain types of drug and health checks for non-exempt employees that don't apply to exempt employees. These types of differential filtering steps can create barriers to employment that only apply to candidates without certain classifications.
These were just the few issues that we have run into, I'm sure there are a lot more that we haven't yet. One important note is that I don't think these problems are issues with the *people* in human resources. We have had lots of good, collaborative relationships with human resources staff at a variety of institutions and still run into these barriers. I think they are due to antiquated *systems* that prevent a modern approach to hiring and throw up barriers, particularly for non-traditional candidates. Sometimes these systems are built for a reason - the FLSA is important and re-classification of employees should not be taken lightly, particularly with respect to employees vs. contractors.
But if we are really going to expand access to modern technical jobs a significant re-think of human resources practices is required. I'd love to talk to people more about this since it is a key challenge for us in helping our program completers as they move into modern jobs.