Sweden

The two percentage-point gaps for Sweden are shown next to the unweighted mean across settings. Both are defined as female minus male. A negative choice gap means that among students who have already cleared the same academic bar, women are less likely than men to rank a STEM program first.

The table summarizes features of the admission system that bear on the construction and interpretation of the analysis. Untranscribed values are marked as pending.

Postsecondary education in Sweden is tuition-free, and all students are eligible for a monthly stipend as well as a subsidized loan. Admission to tertiary education is centrally managed by the Swedish Council for Higher Education (UHR). Students are matched to tertiary education places using a deferred acceptance algorithm that takes into account their ranked preferences, their high school GPA, and/or their score on the university entrance examination, the Hogskoleprovet. As in Finland, but unlike the college admission exams of the other countries in the study, the entrance exam is voluntary.

The system operates multiple parallel admission groups, and students can participate in all of them simultaneously. For each program, at least a third of the vacancies are reserved for a group that competes only on high school GPA. At least another third is allocated based only on the university admission exam. The remaining third of vacancies is also mostly assigned by high school GPA, but can sometimes be used for special admission paths: in some highly selective majors, an additional test or an interview is used to allocate this last third. Admissions through such special groups are not included in the paper's analysis. An applicant's best ranking across the groups in which they compete determines their admission status.

An applicant may rank up to 20 programs, chosen from roughly 15,374 programs offered in a typical year across the 41 institutions inside the centralized system. Places are assigned in two rounds each year, described in the application calendar below.

Two rounds of applications are organized each year: a larger one in April for programs and courses starting in August, and one in October for those starting in January. Students can apply to full-time programs and short courses in the same application. A student can never be admitted to multiple programs in the same semester, but could be admitted to both a course and a program, or to multiple courses.

As in Finland, most students submit their ranked preferences before knowing their high school GPA, so the list reflects expected rather than realized performance. After an initial round of allocations, applicants can choose to accept any offer they receive or participate in the second round for admission to higher-ranked alternatives. Their scores and lists of preferences do not change between the two rounds, but the admission cutoffs might. If a first-round offer is declined, it cannot be recovered.

Every institution in this system admits through the centralized platform: 41 of 41 institutions participate. Sweden is one of only three settings in the study, alongside Australia and Greece, where centralized admissions cover every university, so there is no decentralized sector or outside option to document. This is a strength of the setting: any student who wants a university seat must appear on the platform, so covering the institutional distribution implies covering the student distribution.

Students apply to specific college-major combinations, with roughly 15,374 programs offered in a typical year. No participating institution charges tuition, and financial aid is available to all students in the form of a monthly stipend and a subsidized loan.

The performance measure is upper-secondary GPA and/or the score on the voluntary scholastic aptitude test, the Hogskoleprovet. Because the system runs parallel admission groups, an applicant can compete on grades, on the exam score, or on both at once, and the best ranking across groups determines admission status.

For the analysis, performance groups are formed within setting and cohort over the full distribution of score holders. The measure is constructed identically for women and men; no admission adjustment, weighting, or bonus enters it. In the 2017 cycle there are 108,237 score holders, and the top decile contains 10,758 students: 7,085 women and 3,673 men, so women are 65.9 percent of the high-achieving group against 57.4 percent of all applicants. Women are over-represented at the top of this distribution relative to parity, so the pipeline gap runs in women's favor in this setting.

The manuscript records no setting-specific caveat for Sweden: the performance groups fill cleanly and every field is reported.

The data cover the 2008 to 2017 admission cycles; the analysis sample uses the 2017 cycle. In that cycle the full distribution against which performance groups are defined comprises 108,237 score holders, of whom 10,758 fall in the top decile. Women are 57.4 percent of all applicants in this setting, and roughly 108 thousand students apply per year, of whom about 13.6 thousand are admitted to STEM programs.

Admissions through special groups that use an additional test or interview are excluded. Further record-level exclusions, repeat-applicant rules, and field mappings are documented in the replication code.

Admission rules in Sweden do not differ by sex: the paper's Table A.III comparison of the ten systems records no sex-conditional admission rule for this setting. As the paper's Data section notes for these systems, eligibility runs through academic performance alone, here high school grades and the voluntary entrance exam, and the performance measure used in the analysis is constructed identically for women and men, with no adjustment, weighting, or bonus entering it.

The data come from the Swedish Council for Higher Education (UHR), the agency that centrally manages admission to tertiary education in Sweden, and cover the years 2008 to 2017; the analysis sample uses the 2017 admission cycle. The records contain information on students' high school GPA, their scores on the college admission exam, and their rank of applications. Individual records are matched to data from Statistics Sweden to obtain information on gender, age, and parental education.

Access conditions and the agency contact route have not yet been supplied for publication.