The QS World University Rankings aims to guide students and their families in choosing higher education institutions. Its methodology is based on: an academic reputation survey (40%); an employer opinion survey (10%); number of students per full-time equivalent academic staff (FTE) (20%); number of citations per FTE staff normalized by area of knowledge and year of publication (20%); proportion of international students (5%); and proportion of foreign academic staff (5%).
It should be noted that Quacquarelli Symonds is an educational consulting firm that provides student counselling and holds student recruitment fairs worldwide. The result in this ranking depends more on institutional reputation (50%) than any other similar ranking and, therefore, tends to value the recognition of sectors of society more than the performance of institutions.
Reputation scores consider a continuous weighted average of the last five years. Answers collected in the period from November 2020 to January 2021 receive 100% consideration, the previous year 80%, the previous year 60%, and so on. Citations are measured by the number of full-time equivalent (FTE) professors, normalized by area of knowledge and year of publication. Due to the unreliability of citation analyses for recent publications, the period 2015-2019 is considered in this ranking. Therefore, it is not a reflection of current conditions, but of the recent past.
As usual, we would highlight the importance of the employer reputation aspect of this ranking, as it is the largest survey of global employers of university opinion. This indicator has a number of methodological issues. Employer reputation surveys display very strong Matthew effects because employers lack privileged information. Therefore they tend to vote for the best known institution. Employers are not evenly distributed geographically or by industry. They often do not have close ties to the universities they are passing judgment on. Notwithstanding, it is an important measure for universities to see how external stakeholders perceive them.
Employer reputation scores
The employer reputation scores are based on a combination of domestic and international respondents to QS’s global employer reputation survey. This means that some international employers’ opinions are taken into account. The weight given to the small number of domestic respondents for the survey has a much larger effect on performance in this indicator. This is especially true for universities not considered the best known in their country. People from outside of higher education but in positions of authority within business, government or third sector are invited to give a list of who they think are the best institutions for hiring graduates. They are asked to identify a region with which they are familiar. They are then asked to name up to 10 domestic institutions, and up to 30 regional institutions.
It is important to note the nature of this indicator. Namely, it is not a performance indicator, and does not relate directly to any single activity of the university. It is a proxy measure of how the public think of the performance of higher education institutions.
We took the top 2 ranked universities for a series of countries to assess the relationship between national survey respondent behaviour and employer reputation performance.
QS 2022 Ranking result for Employer Reputation
| University 1 | University 2 | Ave. | Number of universities listed | Survey % (2020) | |
| Brazil | 71.3 | 34.3 | 52.8 | 27 | 1.66 |
| Argentina | 93.4 | 45.8 | 69.6 | 24 | 1.3 |
| Chile | 97.6 | 92.3 | 94.95 | 20 | 1.29 |
| Russia | 76.5 | 48.7 | 62.6 | 48 | 2.61 |
| Colombia | 95.2 | 94.2 | 94.7 | 19 | 1.37 |
| Mexico | 93.2 | 90.4 | 91.8 | 24 | 1.4 |
| Malaysia | 91.7 | 56.4 | 74.05 | 22 | 1.36 |
| Spain | 36.5 | 56.7 | 46.6 | 29 | 2.73 |
| Portugal | 27.7 | 20.2 | 23.95 | 7 | 1.35 |
| Italy | 81.1 | 46.6 | 63.85 | 41 | 2.38 |
| Turkey | 27.5 | 20 | 23.75 | 22 | 1.54 |
In previous technical notes, we observed that the number of universities taking votes per country was perhaps to blame for Brazil’s low score in this indicator. The table above, however, suggests that this is not the case – the number of universities ranked per country has no correlation at all with employer reputation score. Indeed, the country with the lowest scores of all, Portugal, has the fewest universities, while Russia and Spain, who have the most universities, did not perform well either.
Neither is it clear that Colombian, Mexican and Chilean institutions are higher performing than Brazilian, Spanish or Portuguese institutions in any significant senses, and certainly not that would explain a difference of 70 points between Portugal and Chile.
Legal status of the university did not seem to have a marked effect on the employer reputation score either – UNAM scores higher than Monterrey Tech, PUC-Chile and Universidad de Chile’s scores are close to one another, and there is more similarity between them than there is similarity between them and others of the same status. The same is true of Colombia, where Universidad de los Andes and Universidad Nacional are close in score, and closer than either is to their public or private counterparts in other countries.
We have also suggested that the percentage of national respondents had an effect, but in this sample it does not seem to have been the determining factor behind performance either. Spain has the highest rate of responses among the sample, but one of the lowest scores. Chile and Colombia have among the lowest response rates and rank the highest, while their participation rate is almost identical to Portugal, which is the lowest of all. Because there are only sample breakdowns available up to 2020, we recreated the table for that edition of the ranking and noted very little difference in performance.
Linguistic bias
One possible effect for Brazil being lower is that when employers in Latin America are asked to name institutions, they are likely to look to Spanish speaking leading universities as hiring options than Portuguese speaking ones. Therefore, it is more likely that a Chilean respondent will choose UBA or Universidad de los Andes than USP or Unicamp because of the linguistic limitations involved in hiring someone whose first language is not Spanish. The Spanish and Portuguese respondents likewise are probably inhibited by being in an exceptionally competitive local region (Western Europe), where there are a number of powerful institutions who teach predominantly in English, and are more likely to be voted by international respondents.
QS 2020 Ranking result for Employer Reputation
| University 1 | University 2 | Ave. | Number of universities listed | Survey % (2020) | |
| Brazil | 73.3 | 34.5 | 53.9 | 27 | 1.66 |
| Argentina | 91.3 | 44.3 | 67.8 | 24 | 1.3 |
| Chile | 95.5 | 90.8 | 93.15 | 20 | 1.29 |
| Russia | 84.9 | 26.4 | 55.65 | 48 | 2.61 |
| Colombia | 87.9 | 89.7 | 88.8 | 19 | 1.37 |
| Mexico | 91 | 88.9 | 89.95 | 24 | 1.4 |
| Malaysia | 72 | 44.4 | 58.2 | 22 | 1.36 |
| Spain | 45.4 | 38.4 | 41.9 | 29 | 2.73 |
| Portugal | 28.1 | 24.2 | 26.15 | 7 | 1.35 |
| Italy | 84.3 | 46.3 | 65.3 | 41 | 2.38 |
| Turkey | 27.5 | 20 | 23.75 | 22 | 1.54 |
It seems therefore, that the outcomes of the employer reputation survey for flagship universities is based more on social and political environments and relationships that the external community has with its university system as a whole than the performance of individual institutions.
In Brazil’s case, it cannot be coincidence that the score for its two flagship institutions tumbled (by over 10 points for USP and 28 points for Unicamp) in the 2019 edition, launched in 2018. The federal government’s hostile stance and constant impugnment of universities, which in turn led to a series of high profile articles and debates about how higher education has failed, especially in its provision of skilled labour to the market. Such a political movement did not happen in Chile, Argentina, Colombia or Mexico. As a consequence, Brazil’s performance in the indicator diverged drastically from this group.
Given the banded nature of performance in this indicator, we would argue that a more useful way of viewing this indicator is that it is a reflection of general esteem of higher education by a sector of society. As such, it is heavily influenced by public communication and political environments, and should be seen as a collective challenge by university systems, not a reflection of a single institution.
Building knowledge and understanding of the performance of higher education must be undertaken in the state of São Paulo as a collective action in order to have a positive impact for any individual institution.
Priority Actions for São Paulo state institutions
- Create consistent and easy to understand indicators on graduate outcomes and economic impact that are used by all institutions.
- Improve alumni follow-up and recommend to respond employer’s surveys.
- Find international benchmarking for the performance of São Paulo institutions and their contributions to the local economy and graduate outcomes.
- Communicate this impact jointly, showing not just individual institutional excellence, but systemic excellence.
- Engage not just with mainstream media, but also with trade media and organisations in the state.
- Track graduates overseas, especially those working in Latin America, and ensure that they also stay engaged in the university to create a an appropriate perception internationally.
USP Performance
| University | Year | Position | Score | Citations | International students | International Faculty | Faculty student | Employer Reputation | Academic Reputation |
| USP | 2022 | 121 | 55.9 | 42.3 | 3 | 2.4 | 18.8 | 71.3 | 90.1 |
| USP | 2021 | 115 | 54.8 | 38.1 | 3.1 | 8.2 | 20.5 | 67.4 | 88.6 |
| USP | 2020 | 116 | 55.5 | 35.2 | 3.7 | 8.9 | 25.2 | 73.3 | 88.3 |
| USP | 2019 | 118 | 55.5 | 30.7 | 4.3 | 28.2 | 80.4 | 87.2 | |
| USP | 2018 | 121 | 59.1 | 25.7 | 9.6 | 31.4 | 90.2 | 94.8 |
USP is occupying the same position as it did in 2018, although its indicator scores have fallen over the past five years, almost entirely in the Faculty student and employer reputation indicators. The citations indicator has risen for the fifth consecutive year. Internationalisation remains a significant obstacle – for “international faculty”, the numbers are low, and unlikely to advance in the next few years, even if it does become possible to hire new researchers. The devaluation of the real means that attracting foreign researchers on competitive salary packages will be very challenging. The economic situation of the country makes the prospect unlikely to improve. The university’s best prospect for growth in the coming years is attending to the employer reputation survey (see comment above).
Unicamp
| University | Year | Position | Score | Citations | International students | International Faculty | Faculty student | Employer Reputation | Academic Reputation |
| Unicamp | 2022 | 219 | 41.3 | 43.4 | 3.5 | 2.6 | 13 | 34.3 | 65.3 |
| Unicamp | 2021 | 233 | 39.2 | 35.6 | 3.7 | 8.8 | 16.5 | 29.6 | 62.3 |
| Unicamp | 2020 | 214 | 42.1 | 32.7 | 4.3 | 9.9 | 21.1 | 34.5 | 67.5 |
| Unicamp | 2019 | 204 | 43.3 | 31 | 5.1 | 11.1 | 27 | 45.2 | 65.6 |
| Unicamp | 2018 | 182 | 49.5 | 30 | 73.4 | 76.9 |
Unicamp has still not re-entered the top 200, despite improvements in the citations score for five consecutive years. The reason for the university sitting outside this category lies in the fall in reputation scores, and it is in these dimensions that the university has the best chance of improving. That the faculty-student ratio is so low is interesting. Essentially for this indicator, Latin American institutions are faced with a choice – UBA reports all of its academic staff, regardless of whether they are engaged in undergraduate teaching. As a result, the citations score for the university is very low, as large numbers of staff engaged in the licensure process and teacher training are included. Given that smaller classroom sizes are not a strategic priority for Unicamp, maintaining the data as it is to reflect research excellence is the better representation.
UNESP
| University | Year | Position | Score | Citations | International students | International Faculty | Faculty student | Employer Reputation | Academic Reputation |
| Unesp | 2022 | 492 | 24.2 | 15.5 | 4 | 1.9 | 25.3 | 21.9 | 33.7 |
| Unesp | 2021 | 493 | 24.2 | 16.2 | 2.1 | 5.3 | 32 | 16.9 | 30.8 |
| Unesp | 2020 | 482 | 24.6 | 14.7 | 2.2 | 5.9 | 35.5 | 15.5 | 31.2 |
| Unesp | 2019 | 491 | 23.9 | 12.9 | 5.1 | 6.8 | 38.2 | 15.3 | 28.8 |
| Unesp | 2018 | 491-500 | 27.4 | 37.2 |
Unesp’s employer reputation and academic reputation have actually increased slightly over the past five years, but the citation scores remain much lower, and have not grown over the past five years. This is despite the university’s output not being significantly lower in FWCI than USP (Unesp’s is 0.93, USP’s is 1.18). For Unesp, ensuring that it focuses on its areas of outstanding performance, such as animal and veterinary sciences and pharmacology. These areas should be supported as much as possible, and crucially, given support to travel and collaborate abroad in order to gain reputation for the university.