2018-2020 APSA Graduate Placement Survey

The Graduate Placement Survey run by APSA has collected and analyzed data on how PhD candidates in political science perform on the job market for decades. APSA’s annual Graduate Placement Survey examines characteristics of candidates in political science doctoral programs who are entering the job market, recording outcomes of candidates’ efforts to find their first employment. The Graduate Placement survey has been fielded since 2009 in its current format. More recently, APSA has collected data on those entering doctoral programs, or cohorts of incoming students. For the third and fourth years, the survey collected information on characteristics of incoming doctoral students, and their levels and years of funding. This report details our findings for the 2018- 2019 and 2019-2020 academic years, which includes the start of the COVID-19 pandemic in Spring 2020, but does not yet encompass the pandemic and resulting recession’s impact. Abstract. The incoming cohorts of PhD students before the pandemic revealed incremental advancement toward a more diverse profession both in the students accepted and in the allocation of funding, the strongest predictor of placement. The 2019-2020 cohort of incoming students was more gender balanced, and women were more likely to receive full funding than men. The rankings and type of institutions students attend, and their own gender, race/ ethnicity, and/ or home country correlate with variations in the levels of funding received, though students generally receive full funding and/or funding for 5+ years. In 2019-2020, incoming doctoral students from other countries were more diverse in terms of gender/ race-ethnicity than those from the United States. While more diverse students attended public institutions than private ones (as opposed to 2018-2019), URMs/ international students at private institutions were more likely to receive full funding than those at public ones.

As there are only three consecutive years of data on incoming students, consistencies and disparities in the data sets are unable to be isolated as trends. In addition, data is subject to variation in institutional response.
The data analyzed in this report are not weighted. The totals reported here for each survey are the complete number of students reported by departments in the years 2018-2019 and 2019-2020 and may not represent the complete universe of incoming students. We estimate to capture most of the candidates on the job market in normal years. In addition to data collected through the Graduate Placement Survey, we use the National Research Council's rankings of institutions offering degrees in political science, which we aggregate into quintiles.

Abstract.
The incoming cohorts of PhD students before the pandemic revealed incremental advancement toward a more diverse profession both in the students accepted and in the allocation of funding, the strongest predictor of placement. The 2019-2020 cohort of incoming students was more gender balanced, and women were more likely to receive full funding than men. The rankings and type of institutions students attend, and their own gender, race/ ethnicity, and/ or home country correlate with variations in the levels of funding received, though students generally receive full funding and/or funding for 5+ years. In 2019-2020, incoming doctoral students from other countries were more diverse in terms of gender/ raceethnicity than those from the United States. While more diverse students attended public institutions than private ones (as opposed to 2018-2019), URMs/ international students at private institutions were more likely to receive full funding than those at public ones.

I. Characteristics of Incoming Students. *
We examine several characteristics of incoming students, their backgrounds, and institutions, including: gender (female, male, other, not reported), race and ethnicity, home country (United States or international), institutional type (private or public), and undergraduate major (humanities; natural sciences, technology, engineering, math; political science or government; other social science; or other). In order to better understand the changing field of political science, as well as to track cohorts over time to when they enter the job market seeking their first placement, we look at both incoming and outgoing students. We compare the number of incoming doctoral students in political science, as reported by 102 (in 2018-2019) and 44 (in 2019-2020) graduate departments, to the as of yet undetermined number of graduate students reported by those institutions as on the job market. † Gender. The incoming class of political science doctoral students and the cohort of political science candidates on the job market were slightly more balanced in terms of gender in 2019-2020 than previous years. Around 42% of the reported incoming cohort of students in 2018-2019 were women, compared to 44% in 2019-2020 . A similar trend was observed for candidates on the market, where 45% were women (36% in the previous year.) 53% of the incoming students for the 2019-2020 class were men, as were 54% of those candidates on the market, compared to 57% and 63% (respectively) for 2018-2019.
Race and Ethnicity. While most incoming doctoral students for 2018-2019 and 2019-2020 were White/ Euro-American (54%, 52%), followed by East Asian/Asian American (16%,14%), and African American/Black (6%, 8%) students, incoming doctoral students were more racially and ethnically diverse than the student candidates on the job market in the same years. 75.5% of those on-the-job market (compared to 52% of incoming students) were White/ Euro-American.
In 2019-2020 incoming students were twice as likely to be African American/Black -8% of incoming compared to 4% of those on market -possibly indicating less retention of African American/ Black students, proportionally, or changes in recruitment efforts. As more longitudinal data becomes available, we can determine if it indicates differences in retention and recruitment of minority students of significance over time, or if the differences are a result of other factors, such as who goes on the market when and for how long.
Home Country. 57.1% of incoming students for 2018-2019 were from the United States, and 35.5% were International students. For the incoming class of 2019-20, 62.2% of students were from the United States, and 35.5% were International. In 2018-19 the home country of 7.4% of incoming students was unreported, compared to 2.3% in 2019-20. Incoming international students were slightly more gender balanced than incoming students from the U.S. In, both years, more women were international students.
International students were more racially and/or ethnically diverse than the incoming class of U.S students overall. Just 18.3% of incoming international students in 2019-2020 were White, compared to 71.2% of U.S based students. The previous class had a similar composition. 19.5% of international students were White, compared to 77.1% of U.S based students.
The greatest proportion of international students in both years were East Asian/ Asian American (31.8%, 34.4%), a contrast to the number of American students who were East Asian/ Asian American (4.5%, 3.1%). In addition, contrary to the previous year, U.S students were slightly more likely to be Hispanic/Latino/a than international students (11.7% compared to 7.9%). The incoming class of US students was more likely to receive funding for 5 to 6 years in 2019-2020 than international students. However, both groups received full funding at similar rates from 2018 to 2020.
Institution Type. We identified a few variations when comparing the cohort of incoming students from private universities with those from public universities. In contrast to the previous year, 2019-2020 incoming students from public universities tended to be more racially/ethnically diverse (33.2% underrepresented minorities) and more international (37.6% ). However, differences in funding levels to specific groups are evident, as detailed below in this report. Of the 2018-2019 incoming doctoral cohort captured by our survey, 32.9 % were incoming candidates at institutions in the first quintile of the NRC rankings. In 2019-2020, the first quintile covered only 18.3% of reported incoming students . The quintiles for both years show an inverse trend with 2018-2019 having successively less students while 2019-2020 had more students, except for the 3 rd quintile, which contained the smallest proportion of the cohort of incoming students. Distributions of doctoral students across NRC quintiles varied across demographic categories, as will be discussed later on.  While female and male students were roughly equally distributed in the 1 st , 2 nd and 4 th quintiles, women were more concentrated than men in the 3rd quintile, and men were more likely to attend 5th quintile institutions. Women and men were equally likely to attend Unranked institutions in 2018-2019, however more men attended Unranked institutions in 2019-2020.

NRC
Underrepresented minorities were more likely to attend 1st quintile institutions (25%) in 2019-2020, than non-Underrepresented groups 18.20% (compared to 34.4% and 29.4% in 2018). However, they were also more likely to attend lower rank institutions for both years (5 th quintile institutions 18%, 25%). Specific racial and ethnic groups had their own unique distributions across quintiles.
Notably, a large proportion of East Asian/Asian American, Hispanic/Latino/a, Native American/ AK. Native, as well as those who identified their race or ethnicity as Other, were located at 1st or 2nd quintile institutions. In some cases, this notable amount is due to small population size. For example, students identifying as Native American/AK Native were only 0.4% of the incoming student population in 2019, compared to 0.2% in 2018.  (2015)(2016) found a statistically significant relationship between whether or not a student received full funding and whether they received an academic placement, signaling funding status at the start of a program is an indicator of receiving a placement upon program completion.
Analysis of the 2017-18 survey echoed these results, showing that more than 9 out 10 students who successfully made it through their programs and onto the market had full funding (91%), with just slightly more than that placing after advancing onto the job market (92%). The data from 2018-19 and 2019-20 follows the same trend, with the majority of students who received full funding receiving placement after entering the job market.
Where there are disparities in funding at the program start for specific minority groups, there are longterm risks in outcomes that would exacerbate disparities. These include being less likely to complete programs and/ or to receive placements in the future, potentially creating higher debt burdens that are even more difficult to pay off, particularly if a lack of funding contributes to failure to complete the degree, and if degree incompletion is a contributing factor to unemployment, under-employment, or lower pay. For this reason, we examine funding status by race and ethnicity, and by institution type.
It is important to reiterate moving forward, that when examining racial and ethnic minority groups, especially within other variables, we are looking at small populations. As such, the descriptive statistics discussed in this report are susceptible to general variation in institutional response and slight changes in the population. Funding by Gender. For the incoming class of 2018-2019, women and man received full funding at similar rates. In 2019-2020, women (92.3%) incoming doctoral students were more likely to receive full funding than men (82.9%). However, men were more likely to receive partial funding by 6.2%. The same trend was observed between gender groups within international students.
Funding by Race and Ethnicity. Across all races and ethnicities, most incoming doctoral students received full funding. Most groups reflected the same proportions for funding status as the entire population did overall, with the notable exception of Hispanic/Latino/a students in 2019. A greater proportion of Hispanic/Latino/a students were less likely to receive full funding in 2019, just over 78%, the lowest of any racial and/or ethnic group. Contrary to previous years, some groups were more likely to receive full-funding in 2019-2020, with over 95% of African American/ Black students and over 89% of Middle Eastern/ Arab American students received full-funding, compared to 88.6% and 87.5% respectively in 2018-2019.
Incoming students not belonging to underrepresented minority (Non-URM) groups within the discipline (White/Euro American and East Asian/ Asian Americans) were just slightly more likely to receive full funding than underrepresented minority (URM) students in 2018 (92.7% and 91.8%). A shift was observed in 2019, with 88.9% (URM) students receiving full funding compared to 83.5% of (Non-URM) students. More non-underrepresented minorities received partial funding or no funding in 2019.
Disparities in funding by race and ethnicity are more noticeable among men and women. Of any group, men of underrepresented minority groups were the least likely to receive full funding, though they still received funding at high rates (88.2% in 2018, 78.9% in 2019). Conversely, females of underrepresented minorities received full funding over 96% of the time in both years, the highest of any group and around 8.1% higher than their male, URM, counterparts. However, when examining specific racial and ethnic groups within gender, we found a shift in the incoming class of 2018-2019 and 2019-2020. African American/ Black females were least likely to receive full funding of any racial and ethnic group during 2018, with just 84.6% receiving full funding. African American/ Black males received full funding more often than their female peers in 2018 (90%), but the inverse was seen in 2019-2020.
Upon further inspection, we found that programs at two institutions offer no funding to any student at program start and heavily influence the funding rates of African American/Black doctoral students. Both institutions have higher than average proportions of African American/ Black students. In addition, 40% of all African American/Black doctoral students attend either of these two institutions. Removing these two institutions from the analysis drastically changes the results, increasing full funding within African American/Black doctoral students to 93.8% in 2018-2019, and 95.2% in 2019-2020.
These results are in concordance with previous reports. The concentration of minority students at specific institutions demonstrates the effect response rate can have on small populations in our data. Small population groups, such as African American/Black students, who make up 5.6% of incoming students in 2018-2019 and 8% in 2019-2020, may be sensitive to changes in institutional response, as well as general variation in incoming students. It also demonstrates the outsized role these institutions play in preparing scholars of color for the profession compared to other peer institutions.
Examining disparities in racial and ethnic groups within the context of students' home countries, we found a few notable differences. First, in 2018 East Asian/Asian American students from the U.S were offered full funding 100% of the time, while their international counterparts received funding only 91.4% of the time. In contrast, Middle Eastern/ Arab American students were less likely to receive funding if they were from the U.S. (75%), while 90% of their international counterparts received full funding. The opposite was observed in 2019-2020, where international East Asian/American students were offered full funding 100% of the time (compared to 80% of U.S based student) and international Middle Eastern/Arab American students were less likely to received full funding compared to their U.S counterparts. African American/Black students from the U.S were offered more full funding in 2019, compared to 87% in 2018, however, their international counterparts stayed within the 80% funding threshold in both 2018 and 2019.
Funding by Institution Type. Private institutions were more likely than public institutions to offer their incoming students full funding by almost 10% in 2018 (7.4% in 2019) and less likely to offer no funding than public institutions. In 2019, 0% of incoming students at private institutions were not offered any funding at all, compared to 6.6% at public institutions.
13.5% -4.8% -6.2% ------South Asian/Indian American 4% 5.2% -6.2% 6.1% -6.1% 0% Two or More Races 13.5% -4.8% -6.2% ------White/ Euro-American -5.5% 2% 2.8% -2.9% 2.9% 0% Within institution type, there were not many noticeable differences in funding by race or ethnicity. Generally, most groups at public institutions were less likely to receive full funding than those starting doctoral programs at private institutions. While in 2018, both men and women received full funding at similar rates at private institutions, in 2019 women were more likely to received full funding at private institutions than men (by 12.5%) For the incoming class of 2018-2019 at private institutions, the average funding gap between men and women was slightly lower than at public ones with both women and men at private institutions receiving funding more often than their counterparts at public institutions.
However, during 2019-2020 the gap widens, with 100% of women receiving full funding compared to 87.5% of men. Underrepresented minorities and non-Underrepresented minorities at both public and private institutions received full funding at rates largely consistent with the overall average for their institution. There was no funding gap between URMs and Non-URMs at private institutions in 2018-2019, with 100% of both groups receiving full funding. In 2019-2020, 91.3% of non-URMs received full funding compared to 100% of URMs. At public institutions, non-URMs were slightly more likely to receive full funding (89.1% compared to 85.3%), while URMs were slightly more likely to receive partial funding (5.9% compared to 2.1%). Conversely, in 2019-2020 URMs were more likely to receive full funding (88.2% compared to 81.6%) and less likely to receive partial funding (2.9% compared to 7.1%).
There were, however, observable differences in funding by race and ethnicity. These differences were especially pronounced within public institutions. 8.9% fewer than average African American/ Black students, 4.2% fewer of Middle Eastern/Arab American students, and 15% fewer of multi-race incoming students received full funding at public institutions in 2018-2019. For the incoming class of 2019-2020, there was an increase in full funding for both African American/Black students and Middle Eastern/Arab Americans. Hispanic/Latino/a incoming students at public institutions were more likely to receive full funding than non-Hispanic/Latino/a in 2018-2019, however the opposite was observed in 2019-2020. At private institutions, nearly all incoming students-with the exception of those who identified their race as "other"-received full funding at average or above average rates.
Funding by NRC Quintile. Students entering upper NRC quintile institutions were more likely to receive full funding than those at lower-ranked or unranked institutions. In 2018, nearly 100% of students entering 1 st , 2 nd and 3 rd quintile institutions received full funding. Similarly, in 2019 more than 90% of those in the 1 st , 2 nd and 3 rd quintile institutions received full funding. Students in the 4 th and 5 th quintiles were less likely to receive full funding at public institutions. Students in the 4 th quintile received 81.7% in 2018 and 81.8% in 2019, and those in 5 th quintile received 80% in 2018 and 77% in 2019. Students attending an unranked institution received full funding 13% more of the time in 2019-2020.

III. Years of Funding for Incoming Students.**
Most candidates spend five to six years in their program before seeking placement, making funding through their fifth or six years an important factor in successfully making it through a program and on to the job market.

Years of Funding by Gender.
While both male and female students were highly likely to be guaranteed 5 years of funding, some small differences exist. Incoming female students were slightly more likely to be guaranteed funding for a longer duration than males. 91.2% of females were guaranteed between 4 and 6 years of funding, compared to 87.2% of men, whereas 6.5% of males were guaranteed between 1 and 3 years, compared to 4.1% of females, the same trend was observed for the incoming class of 2019-2020. Students who identified as transgender or other were more likely to receive guaranteed funding for 5 years for the 2019-2020 class.
Years of Funding by Race and Ethnicity. Underrepresented minorities and non-Underrepresented minorities received 5 or 6 years of funding at similar rates during 2018-2019. However, we found lower than average rates of funding for certain underrepresented minority groups. 54.2% of Middle Eastern/Arab American incoming students were guaranteed 5 or 6 years of funding, lower than 78.9% of students on average. South Asian/Indian American students, African American/Black students, and multi-racial students also received fewer years of funding at above average rates. A similar trend was observed in 2019-2020, Middle Eastern/Arab American students, East Asian/ Asian American, and South Asian/Indian American students receiving 5 or 6 years of guaranteed funding lower than 70.2 % of students on average.
Hispanic/ Latino/a incoming graduate students were more likely to be guaranteed 5 or 6 years of funding than Non-Hispanic/ Latino/a during 2018-2019 and 2019-2020. A similar number of Hispanic/Latino/as and Non-Hispanic/Latino/as were guaranteed between 1 and 4 years.
Years of Funding by NRC Quintile. Nearly 100% of students at 1st or 3rd quintile institutions received 5 or 6 years of guaranteed funding in 2018-2019 and 2019-2020. In 2018-2019 students at 1st quintile institutions more likely to receive 5 or 6 years than other students. In 2019-2020 students at 3rd quintile institutions were more likely to receive 5 or 6 years than other students. Most students at 2nd  institutions received 4 or 5 years of funding, but less than 10% received 6 years of funding in 2018-2019 and 2019-2020. Students at 4th or 5th quintile and unranked institutions were less likely to be guaranteed 5 or 6 years of funding. They were more likely than other students to receive 1-4 years of funding. Students at unranked institutions were also the most likely group to receive no funding in 2018-2019, while in 2019-2020 those in 5 th quintile institutions were more likely to receive no funding.

IV. Incoming Students Over Time. † †
The 2019-2020 incoming political science doctoral class was similar in gender, racial and ethnic composition, as well as in funding status compared to the 2018-2019 incoming cohort of political science doctoral students, with two notable differences. There was a 3.8% drop in students who identified as an 'Other' race/ethnicity, as well as an increase in those students who identify as Transgender. This is in part due to a change in the survey instrument. Previously, Hispanic or Latino/a was included as a separate measure from the race/ ethnicity category, with many of those who identified as Hispanic or Latino/a selecting 'other' for their race/ethnicity. Similarly, another instrument was added to the gender category. For the 2018-2019 academic year, the survey instrument was updated to include Hispanic or Latino/a in the race/ethnicity category, and in 2019-2020 another update was made to include Transgender in the gender category.
For this year's incoming doctoral class, the proportion of U.S based students increased by 5.1%. There was no difference in the number of international students when compared to the previous year.
For the incoming class of 2018-2019, it was found that 91.8% of underrepresented minorities received full funding compared to 93.4% overall, largely driven by the number of East Asian/Asian American (94.10%) and Hispanic/Latino/a (96.80%) students receiving full funding. Though similar in size, gender breakdown, and racial and ethnic composition, the 2019-2020 incoming doctoral class in political science had a different funding composition for URM students. Generally, incoming URM students for the 2019-2020 academic year were more likely to receive full funding than incoming non-URM students. African American/ Black students received full-funding at above average rates (95.20%), they were more likely to receive full-funding than their counterparts in 2018-2019.
Comparing institutions that responded to both the 2018-2019 and 2019-2020 surveys, this difference persists. The 2018-2019 cohort is slightly less diverse, and is more male, as well as proportionally more White/ Euro American than the incoming students in the 2019-2020 academic year. In the 2018-2019 cohort, non-URM students were more likely to receive full funding than their URM counterparts, the opposite was observed for the 2019-2020 cohort. African American/ Black Students received full funding more often in 2019-2020 than in 2018-2019 by a difference of 8.3%. Additionally, incoming Middle Eastern/ Arab American students were still much more likely to receive full funding starting in 2019-2020 (89.5%) compared to previous years.
With only four consecutive years of data, we cannot isolate the factors driving the differences in funding to racial and ethnic groups. Our findings may simply indicate variation between incoming doctoral classes. Additionally, underrepresented minorities make up a small proportion of all students overall (28.6% in 2016-17, 25.5% in 2017-18, 17.2% in 2018-19, and 22.1% in 2019-20). Moreover, specific racial and ethnic groups have even smaller proportions; for example, in 2018-2019, African American/Black students made up just 5.6% of the incoming doctoral students in total.
These small group populations make these proportions more sensitive to changes in the number of students receiving funding or no funding. Adding or losing a few students receiving full funding can significantly alter results. In addition, given that some of these small student group populations tend to cluster at fewer institutions, variation in survey response at the institution level can also alter results.

V. Conclusion.
Examining the trends in our data, a theme becomes clear. While students generally receive full funding from their Political Science PhD program and generally receive funding for at least five years, students receive this in different magnitudes depending on the characteristics of the institutions they attend, their own gender, race/ ethnicity, and/ or home country. These factors in combination also drive alterations to students' levels of funding, whether the funding is full, or partial, or if students have no funding at all.
In the 2019-2020 academic year, incoming international political science doctoral students from other countries were more diverse than those from the U.S. Students who identified as women tended to receive more funding than men. For the class of 2018-2019 students at private institutions were more diverse than at public institutions, while the opposite was observed in 2019-2020. Contrary to previous years, in 2019-2020, underrepresented minorities and international students at private institutions were more likely to receive full funding. While underrepresented minority groups are more likely to receive full funding overall, African American/ Black students, especially male students, are less likely than average to receive full funding, in part because of the tendency for larger proportions of this group to attend the same institutions -institutions that have less funding for students or may not offer any funding to students at all.
Comparing the 2019-2020 cohort of incoming students to that of the previous academic year, or the first year this information was collected, there are a few key differences, even when just examining institutions that responded to the survey in both 2018-2019 and 2019-2020. The 2019-2020 cohort was more gender balanced than the students reported in 2018-2019. While in both years incoming students largely received full funding, as well as funding for at least 5 years, underrepresented minority students received funding at different rates, receiving funding at similar average rate as 2019-2020 compared to lower-than-average funding in 2018-2019. While we expect variation between cohorts, it is not possible to isolate spurious correlations from relationships until we have more longitudinal data on incoming PhD cohorts.