Getting Into Consulting as a Stata, Python User

[Jared Greathouse]

This may not, unfortunately, be the best place to ask such things, but Statalist has helped me a lot from undergrad to grad school, so I might as well ask it here.

As of April, I finished my PHD coursework (in the United States, that means I have only my dissertation and teaching to worry about, as a PHD Candidate). I figured since I now have more time I'd get into independent consulting for data cleaning, advanced causal inference, and research assistance with stats more generally. I suppose my question here is does anyone have experience in this? How far along were you in your education when you started, how did you manage it and make any progress, and what are the main things I may wish to keep in mind for those who are interested in being independent consultants (who still happen to work and be in academia for the long term)? If it helps, I already have a (new!) website (see my bio on my profile) that sort of explains what I do/am interested in. As usual, I'd appreciate any feedback.

[George Ford]

I've used (and paid) a Stata/R coder on upwork.com. Didn't care about his education. He did a great job. Might look there to get a feel for what's available.

I know that A. Shah (of asreg/asdoc fame) does some consulting.

[Erik Ruzek]

I'm glad you posted this, Jared. Congrats on almost being done with your PhD and considering your next steps.

I've dabbled as a stats and methods consultant on and off. I am doing it right now a little more seriously because I left my full-time academic-adjacent industry/non-profit position. I would say that you need to think hard about and define clearly what type of work you want to do. For example, I realize that I generally prefer consulting with academic researchers moreso than with industry or non-profit clients. This is mainly because my methodological skillset and interests are well-suited to academic research, as is my substantive expertise. Simultaneously, I find there to be generally higher data quality in academia, a deeper appreciation for measurement, and a willingness to live with null results. There are exceptions to this, of course.

Having been out of my PhD for nearly 12 years and subsequently working in academia (8 years) and industry/non-profit (4 years), I've made some deep contacts in both areas. So people know my skillset and will often come to me with offers of work. To make a living off this kind of work, I have the impression that you have to be willing to market yourself and your skills and probably take some jobs that are not really in your wheelhouse and/or have less than ideal data.

There are a few full-time consultants who post here and have built a business out of it, including Ariel Linden and Joseph Coveney. They might be able to chime in with some thoughts, which I would also be really interested to hear.

[Dimitriy V. Masterov]

I have done consulting work three times in my career: as a PhD student in economics in place of teaching, after graduating during the financial crisis when there were very few industry jobs, and now for 2.5 years as a fractional economist in tech (after full-time roles at Domino's Pizza, eBay, and Thumbtack).

If your goal is to stay in academia, I would advise focusing on your research publications and putting off consulting until you get tenure. First, the benefits are pretty low until you establish yourself as an expert in a field and the opportunity cost is high. For instance, on data wrangling, you are competing with folks on Upwork, so prices are driven down by international competition to the point where tutoring may be a better opportunity (e.g., https://www.statatutor.com/). Second, many academic advisors regard consulting as a distraction and will be reluctant to invest in you, which hurts your scholarly prospects. Academics seek to maximize the number of intellectual descendants, and consulting is like going into a monastery. This will make it hard to get consulting time and reduces what you can charge post-tenure.

One notable exception is engagements where the consulting can lead to a high-quality research project with high certainty. This is pretty ideal, but also very rare. Make sure you get a formal data-sharing agreement from your university's external research office before doing this.

If you are curious about industry, do a summer internship at a tech company. These are generally well-paid and will give you a taste of that kind of work. Alternatively, do some work for a professor who does consulting work. Comp here tends to be lower, but you won't waste much time on business development.

[George Ford]

If you're in academics, then getting coauthorship for the work is likely more valuable than consulting.

[Jared Greathouse]

Frequently, people reach out to me who, sometimes, lack formalized training in econometrics. I can think of one example (I won't say who they were, but almost everyone was of Ivy League status). They wanted to do inference in a setting where we have multiple treatments and multiple exposed groups across multiple states.

I basically told them they were biting off way more than they (or anyone, really) could chew. I knew that if I agreed to work with them (their paper had already been rejected from a few journals), I would be doing most of the coding and analysis, and I would have to read about.... however we'd even try to do estimation in this scenario. In other words to your point, George Ford, my time is valuable. Since this isn't my paper, and they really just wanted my methods skills, I figured this would be an example of when I'd do consulting. I ultimately didn't (the lead author didn't know how to listen) work with them, though.

When I asked my friend in econ at MIT told me "If the research is something you care about (e.g., econometrics, stats or the substantive matter/topic) then you do it for the authorship. But, if you're being asked to consult on a topic that you don't really care about, then my time is very expensive." My advisor sort of does similar things, he'll consult for a firm (the term he uses, humorously of course, is "hitman"), really just running simple regressions and explaining the written results. He told me he makes 10-15k per project. Even if it were half that, right, that would be perfect, for me.

As Erik Ruzek suggests, "full time" consultants would need to market themselves and such. I also agree on the data quality points and such. My heart is in academia, that's what I wanna do for the very long term. So, I wouldn't wanna go into consulting completely full time, it's mainly just for extra money (and also, of course, before I love this work). I agree, though, that my skillset (econometrics, causal inference) is mainly more marketable to academia than the stuff industry types typically care about. So, I also agree a lot on being very clear about what specifically the areas I'd be interested in consulting for, are. To Dimitriy V. Masterov's suggestion, yeah, I sorta got the idea from my advisor since he does consulting. He's also been quite supportive of my desire to do consulting (as I mentioned above, I even applied for the firm he's consulted independently for, recently).

But I do see that it's a pretty big decision that (could) affect my academic career. If I already didn't have my dissertation pretty well mapped out, I think I'd be more worried about it, in terms of how it affects progress.

[George Ford]

I'd think it would be difficult to charge an academic enough money to make it worth your time, unless it was hooked to large grant.