Demystifying recruitment

Our group’s (Development Economics Research Group, or DECRG) recruitment effort is not over yet, but we’re slowly getting there. There are various aspects of recruitment that I have been wanting to talk about, having overseen a very large recruitment effort for our department, but I want to start by telling you a little bit about female/male ratios in our hiring.

[A couple of caveats. First, we coordinated our hiring with DIME, which has recently become a separate department in DEC, but this piece will mostly refer to the DECRG effort. Second, the reasons behind the heavy male skew at senior levels in economics deserves a separate discussion and I will not get into those here. Legitimate (numbers of female PhD economics students were substantially lower 15-20 years ago) and illegitimate (explicit discrimination in hiring and promotions, females bearing a higher share of the joint location burdens, the culture of economics that has been hostile to females, etc.) reasons have led to the facts about the stock of staff numbers cited below. Some of my colleagues have written about diversity and inclusion at the World Bank and you can get an introduction here.]

For background, we have three grade levels below our director in DECRG: economist, senior economist, and lead economist (grades F, G, and H, respectively). What follows are my counts from our internal website, which are neither official nor exact: as long as they are approximately close to the exact numbers, they will suffice to make the points. Compared with academia, we do quite well at the junior (economist) level, with 13 males and 9 females. The ratio gets worse at the senior level, 12 to 6, and really skewed at the lead level, with 15 to 3. Hence, if we want to reach parity in our department with respect to the female/male ratio, we need to hire more females at the junior and senior levels (budgetary considerations do not currently allow us to recruit at grade H-level).

Now, back to this year’s recruitment effort. We have pursued two avenues: junior hiring (through Job Openings for Economists or JOE) and a smaller effort to hire non-junior researchers. Let’s start with the latter…

For the non-junior recruitment effort, we asked our colleagues in DECRG to nominate researchers for us to approach. The description was “anyone that you’d like to have here as a colleague, received their PhD within the past 3-9 years (more or less) – regardless of whether you think they are on the market or not.” This produced 49 names, 24 females and 25 males, to each of whom I wrote an email to invite them to apply for a position with DECRG: 24 individuals did so, evenly split between male and female. After reviewing packages, we flew out seven of them (between mid-November and X-mas), three females and four males. While the recruitment effort is still ongoing, a summary of these phases so far, disaggregated by sex, is below:

For the junior market, we received 1,052 applications to our JOE ad. Approximately two thirds of these applications were from male candidates. I say “approximately” because JOE allows people to skip the question on sex or for them to choose “I prefer not to say.” More than 400 applicants chose one of these two options and I am too lazy to go through their names and profiles to categorize them (especially because they did not want us to do so). But, of those who chose to indicate male or female, 64% were male. The female skew in the indeterminate group would have to be substantial (about 72%) for the ratios of male and female applicants to be equal. So, I will loosely assume that we started from a highly male-skewed pool of candidates.

We assigned approximately 150 candidates to each of the seven recruitment committee members, who shortlisted the candidates for interviews at the ASSA meetings in San Diego. Each committee member had the difficult job of shortlisting roughly one out of every 15 candidates in their pile of application packages. This meant that it should not have been hard for each of the members to generate a shortlist that is more or less balanced between males and females, a desire that we did communicate to the committee members. As we are trying to hire a large number of people this year (about 6-8 is our current goal), we shortlisted 78 candidates for these interviews: 36 females and 42 males. So, we did unwind a fair amount of the male skew through our shortlisting effort (from roughly two thirds in the application pool to 54% in the interview shortlist), but not entirely. Part of the difficulty is that it is still hard to find female candidates in certain sub-fields, such as macro, infrastructure, and transport (a little more on this below).

After interviewing these candidates in San Diego, we invited 24 of them for visits to our department (flyouts). 14 of these flyout invitations went to females and 10 to males, flipping the original ratio of applicants. Of these invitations, all of which were scheduled between January 13 and February 21 (filling almost every business day and giving us a marathon of really high-quality seminars), four people cancelled their visits because they received offers that were preferable to DECRG: three of these candidates were female and only one male. We lost our only two female candidates in macro this way, reinforcing the difficulty of filling certain positions with female candidates, as the demand currently seems to outstrip the supply. In addition, there were a few candidates who accepted other positions elsewhere before we had a chance to make them offers. A summary of the junior market effort (so far) is below. One thing to note is that some candidates get offers after others decline, so the equilibrium number of offers does not provide the entire picture:

So, this is how more than 1,100 junior and non-junior candidates were funneled down into 13 offers so far – four of which were accepted (3F/1M), six were declined (3F/3M), and three of which are still outstanding (1F/2M), ensuring parity (in this one aspect) for this year’s cohort of early career hires. Questions and comments welcome…