Career paths of the International Mathematics Olympiad (IMO) medalists

Abstract

Each year, six exceptionally talented high school students from each country participate in the International Mathematics Olympiad (IMO). We thoroughly searched the internet to find the career paths of 2785 IMO medalists who participated in the competition from 1986 to 2005. We have found that the career paths of IMO medalists depend on their home country, despite their similar levels of mathematical talent. For example, the extent to which medalists studied abroad varies significantly among countries in the same geographical region and with comparable income levels. We also found that there is a high concentration in medalists’ foreign education. For example, MIT educated more foreign medalists than any country except for the United States both at the undergraduate and Ph.D. levels. Roughly half of the medalists ended up working abroad. Education is found to be an important channel for the migrant medalists. Although majority of the medalists were educated in mathematics, more than three quarters of their non-academic jobs are either software or economics-finance related jobs. Most medalists who became academics are specialized in mathematics. While the rate of becoming a mathematician academic is not dependent on the income level of the medalists' home country, those from higher-income countries tend to have higher academic performance.

Introduction

Six high school students from each country are selected to participate in the International Mathematics Olympiad (IMO) each year. All participants meet under the same roof to solve six challenging questions over 2 days. Half of them receive a medal signifying a comparable level of mathematical ability to other medalist peers. They compete under equal conditions, regardless of their country of origin. However, due to differences in opportunities, their career prospects vary.

The main question in this paper is whether gifted children with extraordinary academic ability had equal opportunity in their careers depending on their native country. In particular, we wonder whether the income level and geography of the countries affect the opportunities for these gifted children. It is also of interest to explore whether medalists from countries of similar income and geographical location experience similar opportunities. To address this main question, we pose the following specific questions:

• Do medalists receive education at the best global universities at the same rate regardless of their country of origin?

• Which countries have educated the most foreign-educated medalists?

• Do domestically educated medalists have equal opportunities within their country? Are they concentrated in the best universities in their respective countries?

• Do medalists pursue similar fields in their education, particularly in mathematics?

• How does the rate of pursuing Ph.D. education depend on the country of origin?

• Do medalists migrate after their education? Does their education serve as a stepping stone for migration? Which countries experience the most brain drain, and which countries attract the most medalists?

• Do the rates of those who pursue academic jobs depend on the country of origin? Do their publication performances depend on the country of origin?

• Do medalists pursue the same types of non-academic jobs? Are there any global firms that employ a considerable number of medalists?

To answer these questions, we conducted a comprehensive internet search to examine the career paths of medalists. We explored their initial and Ph.D. degrees, as well as their current occupations. Specifically, we assessed the extent to which medalists pursue education abroad and how it relates to their home country. We collected information about the institutions that medalists attend to determine if winning a medal guarantees admission to top global universities. We also analyzed whether their current occupations are related to mathematics education. Lastly, we investigated whether the publication performance of academic medalists is influenced by their home country.

It may be disadvantageous to be born in a low income country, therefore, medalists from these countries have overcome obstacles to stand on equal footing with their peers. If career opportunities were the same for all medalists, we would expect medalists from different countries to have similar education and occupational outcomes. This paper sheds light on whether such an equality of opportunity exists.

The migration of the best talents is problematic for countries seeking to develop industries requiring high-level intellect. Since we documented country-level variation in the migration of IMO medalists, the policy makers looking to attract IMO medalists would benefit from examining the educational policies and cultures of countries that has been successful in retaining most of their top talent.

Related literature

Saul and Vaderlind (2022) provide a comprehensive literature review of IMO competitions. Most studies related to the IMO focus on the competition itself. For instance, their review summarizes numerous papers that analyze the content of IMO questions and the participation and performance of girls. Our study, on the other hand, contributes to a narrower literature by analyzing the post-competition outcomes.

Most reports, papers, and web pages related to the career prospects of IMO medalists focus on a limited number of countries. For example, Zhou and Liu (2009) exclusively examine the career paths of Chinese medalists. Gibson and McKenzie (2011) investigate the migration of 1800 exceptional students who are members of mathematics and chemistry olympiad teams, valedictorians in prestigious high schools or best achievers in national-level college scholarship exams from New Zealand, Tonga and Papua New Guinea.

Among single-country studies Campbell and Walberg (2010) is closest to our study. They examined the career outcomes of US Chemistry, Physics, and Mathematics olympiad participants, identifying the universities they attend, their current occupations, and their publication performances. Our study complements those that include few countries by enabling inter-country comparisons through a larger sample.

There are two related papers that involve extensive efforts to collect career paths of the medalists from all countries (Agarwal & Gaule, 2020; Agarwal et al., 2023). Their data-collection effort largely relies on medalists’ official academic pages and Linkedin profiles. In contrast, our data is obtained through a more comprehensive internet search. While both datasets cover a span of 20 years, their data starts 5 years earlier than ours. Due to these factors, they were able to find information about the current occupations of only 55% of the medalists, whereas we have current occupation data for 84% of the medalists.

Despite our extended coverage, we have replicated some of their results. For instance, like Agarwal and Gaule (2020), we found that academics from higher-income countries tend to be more successful. However, some of our findings differ from theirs. Notably, the migration rate among medalists in our study is significantly higher than the rate found in Agarwal et al. (2023). Additionally, we have gathered data on the first university degree of the medalists, which is a valuable source for understanding country-level differences. While their studies rely on regression analysis, our aim is to show inter-country differences, so we use descriptive statistics to present country-level information.

IMO medalists possess extraordinary academic ability, categorizing them as gifted children. Consequently, our analysis is connected to studies exploring the career paths of the gifted individuals. For instance, Gross (2006) investigated the impact of grade skipping on the career trajectories of gifted Australian children with an IQ over 160. The study concluded that skipping grade places gifted children in a more suitable educational environment, leading to improved academic performance. Additionally, Ozcan (2017) conducted interviews with gifted children to understand how they make decisions regarding their future career choices.

Our study is a quantitative analysis that involves counting the number of medalists in various occupations paths. In contrast, there are qualitative studies that survey Olympians to understand the reasons behind their career choices. For instance, Lengfelder and Heller (2002) sought to understand why female Olympiad trainees struggle to make the national team, identifying trainers’ bias towards males as the main reason for underrepresentation. Jung and Lee (2021) interview 15 former Australian Olympiad medalists to explore how they chose their career paths. The study revealed the positive motivational affect of Olympiad training, particularly in pursuing mathematics during university education. Furthermore, Rebholz et al. (2022) provided mathematics Olympiad training to third and fourth graders in Germany. Their study not only observed an improvement in mathematical capability but also identified motivation for studying mathematics generated by the social environment of the training class.

Our paper is also related to brain drain studies. Docquier et al. (2007) find that the small, poor and politically unstable countries are more likely to experience the emigration of skilled labor. In traditional brain drain studies, all college graduates are typically considered skilled. However, this study focuses on individuals with exceptional skills. The career opportunities for this exceptional talented group differ significantly from those of an average college graduate, leading to distinct career paths.

We examine the career paths of individuals who have achieved early success in their careers, making this a forward-looking study. In contrast, there are backward-looking studies that analyze the backgrounds of those who have achieved success at a later stage. For example, Ciftci et al. (2023) and Yuret (2017) investigate the academic backgrounds of successful Turkish and US academics, respectively.

Data

The list of IMO participants and their performance in the exam was obtained from IMO’s official page.¹ We conducted a comprehensive internet search for the career paths of 2785 participants from 89 countries who received a medal in an IMO competition between 1986 and 2005.² Specifically, we collected information on the medalists’ university degrees and current occupation information as of mid-2023. For some medalists, finding the information was easy, as they had an academic page or Linkedin profile that indicated their medalist status. For others, we conducted a more thorough search to obtain their information. We used the official pages of the national IMO teams, reports, news, social media platforms (e.g. Facebook), and other internet sources that locate IMO medalists. We also conducted searches in local languages and explored possible name variants if the IMO medalist’s name had been converted from a non-Latin alphabet.

We had access to the full name and high school graduation year (i.e., the last IMO participation year) of the medalists, and we matched this information with the internet profiles that we found. However, name confusions are inevitable. For instance, consider a scenario in which there is a person with the same name and high school graduation year as the medalist, and this individual has had a successful academic career, while the medalist leaves no trace on the internet. In such cases, we might have mistakenly selected the wrong person. While we consider this scenario highly unlikely, it is important to interpret our results with this potential bias in mind.

We were able to find information about either education or career for 89% of all the medalists. We have career information for 84% of all the medalists. The rate for missing information varies from country to country. We have information about all medalists from 23 out of the 89 countries. On the other extreme, we could not find any information about any of the North Korean medalists.

We obtained publication information from Scopus.³ We gathered population and GDP information from the World Bank.⁴ We used United Nation’s geoscheme for categorizing countries into geographic regions.⁵

Career path of medalists

Participation in IMO

Table 1 shows that the countries have different performances in the IMO depending on their income, population, and geographic region. During this period, there were a total of 7630 participations, which half of them resulting in a medal reward. It is important to note that the number of participations differs from the number of participants, as some individuals participated multiple times. As a result, the number of medalists is lower than half of the total participations.

Table 1 Differences in IMO performance by GDP, population, and region of the participants' country

There is a reverse relationship between a country’s income and its performance in the IMO. The same percentage of participations from high and low-income countries receive a medal, but a higher percentage of medalists from low-income countries receive a Gold Medal. It is important to note that these countries are all able to produce at least one medal in the IMO, and there is an equal number of countries in each category. Therefore, low-income countries may not necessarily have very low incomes.⁶ Nevertheless, this outcome highlights the availability of educational opportunities for the most talented individuals in countries with relatively lower incomes.

Participants from more populated countries tend to perform better in the IMO competitions. This is expected, as more populated countries have a larger pool of students who may possess talent. There are notable disparities among different geographical regions. Participants from North America, Eastern Asia, and Eastern Europe tend to outperform those from other regions. It is worth noting that communist countries placed significant importance on these competitions. While our data begins at the end of communist era, part of the superior performance may be attributed to the ex-communist countries in Eastern Asia and Eastern Europe. The relatively modest performance of participants from advanced Northern European countries may be due to their emphasis on providing equal opportunities for all children, rather than focusing on selective streaming to identify the best talent.

Medalists’ undergraduate degrees

IMO participants are high school students, and they typically attend university immediately after participating in the competition. Therefore, participation in the IMO is expected to impact their university education opportunities, particularly for their initial university degree. In this section, we examine how these individuals with recognized mathematical talent are distributed among different universities for their first degree.

The initial university degree has different names and structures in various countries, but we uniformly refer to them as undergraduate degrees (UG) in this paper. We were able to obtain information about the UG institutions of 76% of the medalists. It is possible that we did not acquire information from the medalists who attended less prestigious universities disproportionally. Therefore, our results should be interpreted with this caveat in mind.

Table 2 presents the top two institutions from each country that educated most IMO medalists as UG students. The table lays out institutions in 26 countries that educated 25 or more IMO medalists. The count of medalists includes both domestic and international medalists. The highly talented IMO medalists could potentially gain admission to many institutions within each country. However, we observe that the majority of medalists chose to attend just one or two institutions within each country. The only exceptions were Australia and Germany where less than the majority of medalists are educated in two institutions in these countries. In 19 out of 26 countries, the top two institutions educated more than 70% of the medalists.

Table 2 Two UG institutions that educated most IMO medalists

Some countries have a single institution that dominates the education of medalists. The most extreme example is Taiwan, where all medalists earned their UG degrees from the same university. Other countries where a single institution educated more than 80% of UG degree holders include the United Kingdom, Iran, Hungary, Czech Republic, Korea, Bulgaria, and Belarus. It is interesting to note that Cambridge educated majority of the medalists in the United Kingdom even though the country has many other prestigious institutions.

Table 3 reveals that only 487 medalists pursued their UG degrees internationally. Foreign-educated medalists constitute just 17% of all medalists and 23% of those for whom UG institution information is available. Approximately half of the foreign-educated medalists attended universities in the United States. Only seven countries attracted more than ten medalists over the 20-year period we cover. There are ten institutions that attracted more than ten medalists. Five of these institutions are considered to be the top five global universities in mathematics according to USNews,⁷ and they are Stanford, MIT, Princeton, Cambridge and Harvard in descending order. These five universities collectively educated a total of 200 medalists from foreign countries.⁸ It is interesting to note that MIT alone had more foreign-educated medalists than all countries other than the United States. These facts highlight the global inequality in terms of attracting foreign talent.

Table 3 Foreign education (UG)

The low rate of medalists who have foreign UG education indicates that there is little competition for talent at the UG level. Several factors contribute to this outcome. Some countries have strong institutions, so medalists do not need to travel abroad for a quality education. Alternatively, it may be the case that the students who are not proficient in foreign languages face limitations in pursuing international education opportunities. Agarwal et al. (2023) conducted a survey to the IMO medalists and concluded that many medalists aspiring for a foreign UG could not go overseas due to financial constraints.

The inter-country differences in foreign UG education, as presented in Table 4, may shed light on the possible reasons for the low mobility of medalists at the UG level. The second and third columns of the table show the number of medalists and the number of medalists for whom we have information about their UG education, respectively. There is a stark difference in data availability. We know most of the medalists’ UG education for Chinese and US medalists, but we have little information about Vietnamese and French medalists. The fourth column represents the number of foreign-educated medalists, and the fifth column shows the number of medalists who studied abroad in the top five global universities. The last column indicates the ratio of foreign-educated medalists to the medalists for whom we have UG information. Table 4 is limited to 32 countries with at least 25 medalists for whom we have UG information.

Table 4 Mobility for UG degree

We observe that the USA, Australia, UK, France, Israel, and Japan did not have any medalists who obtained a foreign UG degree. These countries are all advanced with prestigious institutions of their own. There seems to have been a lack of competition in attracting global talent among these countries. For instance, no US medalist attended Cambridge, and no medalist from the UK attended MIT. However, not all advanced countries with prestigious institutions have all their medalists earning UG degrees within their home institutions. For example, 20% of Canadian and 79% of Singaporean medalists studied abroad.

The rates of foreign UG education vary among the neighboring countries. Only 10% of Turkish medalists pursued UG education abroad, whereas this rate is 43% for Bulgarian and 41% for Greek medalists. More than half of Romanian medalists obtained their UG degrees from abroad, whereas the rate is only 7% for Hungarian medalists.

It is interesting to note that the number of medalists who went abroad to attend a top 5 global universities also varies significantly between countries. For instance, 24 Romanian medalists attended these universities, whereas no Slovakian medalists did so. Singapore had 24 medalists attending top 5 universities, whereas only 3 Taiwanese medalists did the same.

The differences in foreign education may stem from the strength of domestic undergraduate institutions as well as the lack of opportunities. However, we observe vast differences among countries in the same geographical area with comparable national income. The educational and cultural reasons of the lack of opportunities for the best talents in similar countries are important questions that our data is unable to address.

Figure 1 provides data for medalists’ UG majors. There is a high concentration of choices among the medalists. Specifically, 88% of the medalists studied in mathematics, computer science, and electronic engineering (CS-EE). The medalists clearly possess a comparative advantage in mathematics. However, their preferences could have been different. Only a few medalists have became mechanical engineers or sociologists, even though these fields could also benefit from their high mathematical talent.

Fig. 1

UG Fields of IMO medalists

There are differences among medalists in terms of major choices depending on their home country. Among the 32 countries for which we have UG degree information for at least 25 medalists, less than 40% of the medalists from Belgium, Brazil, India, and Thailand chose to study mathematics, whereas this rate is above 85% for medalists from Romania, Hungary, United Kingdom, Russia, Ukraine, and Germany.

We only have information about the UG majors of 72% of the medalists, so our finding of a high concentration of UG majors should be interpreted accordingly. It is relatively easy to find internet traces of a medalist who has computer science UG degree and has become a software engineer in a global company. However, a medalist who has become a lawyer for domestic courts may not have any internet trace.

Ph.D. degrees of the medalists

We found that 1981 medalists obtained post-graduate degrees. It is important to note that this number represents a lower bound, as we may not be able to find all post-graduate degrees. For instance, we do not have any educational information about 13% of the medalists, so we may not have information about their master’s degrees either. Additionally, some medalists might not have disclosed their master’s degrees if they left their Ph.D. studies at the “All But Dissertation” level. The number of medalist known to have a post-graduate degree is equivalent to 82% of the 2422 medalists for whom we have some educational information, and 71% of all 2785 medalists.

In this section, we focus on the Ph.D. degree because we have more comprehensive information about this post-graduate degree. There is an internet trace for almost every Ph.D. degree earned. 1689 medalists obtained their Ph.D. degrees which is equivalent to 61% of all medalists. The ratio of Ph.D. degrees obtained from abroad to all Ph.D. degrees earned is also 61%.

Table 5 provides statistics for 25 countries, each of which has at least 25 medalists who obtained a Ph.D. degree. We do not see any lack of opportunities for the medalists to obtain Ph.D. degrees. Only Bulgaria had less than half of their medalists earn Ph.D. degrees.⁹ On the high end, Iran and Romania have more than 80% of their medalists earning Ph.D. degrees.

Table 5 Mobility at Ph.D. level

We previously mentioned that many advanced countries had all their medalists educated at home institutions at the UG level. The situation is different for Ph.D. degrees. Only US medalists obtained their PhDs in their home country exclusively. Although the ratio of foreign PhDs is low in some advanced countries such as France and Japan, the ratio is higher than half for other advanced countries such as Australia and Canada. Vast differences exist among countries from the same geographic area and with similar income levels. For example, only 14% of Polish medalists obtained their PhDs abroad, whereas this rate is 98% for Romania. However, the ratio of foreign PhDs in almost all developing countries is high, leading us to conclude that the mobility of the IMO medalists is much higher at the Ph.D. level compared to the UG level in these countries.

Table 6 presents the destination of the medalists who earned their Ph.D. degrees abroad. Among countries that educated most medalists at the Ph.D. level, there is an even starker concentration compared to UG degrees. A total of 68% of the Ph.D. degrees were obtained in the United States. Moreover, 17 out of the 22 institutions from which most medalists earned their foreign Ph.D. degrees are also in the United States. Notably, MIT educated more foreign medalists at the Ph.D. level than any country except the United States. In fact, five US institutions granted more Ph.D. degrees than all of Germany.

Table 6 Destination countries and institutions for foreign Ph.D.

Figure 2 illustrates the medalists’ Ph.D. degree fields. The distribution of Ph.D. fields is similar to that of the UG major distribution. We can conclude that there is limited selection of Ph.D. fields from different UG majors. The ratio of UG degrees in mathematics is 64%, while PhDs in mathematics account for 70%. Conversely, the ratio for UG degrees in CS-EE is 22%, but PhDs in CS-EE make up 17%. The lower ratio of CE-EE at the Ph.D. level is interesting because, as we will see in the next section, many non-academic jobs taken by medalists are computer-related. Although the ratios are very small to draw strong conclusions, we observe that the ratio of economics-finance PhDs is higher than that of UG majors in economics-finance, and the ratio of engineering PhDs is lower than that of UG majors in engineering.

Fig. 2

Ph.D. Fields of IMO medalists

Current location

We have observed that many medalists have pursued education abroad. It is a generally accepted that education is a crucial factor for a country’s growth (Krueger & Lindahl, 2001; Squicciarini & Voigtlander, 2015; Woessmann, 2016). Consequently, the departure of top mathematical talent poses a significant handicap for developing countries. In Table 7, we list 38 countries where we have information regarding the current whereabouts of at least 25 medalists. As seen in the second and third columns of the table, data availability varies from country to country. We have information on the whereabouts of almost all Hungarian and Romanian medalists, but the data availability is very limited for medalists from Hong Kong and Vietnam. The disparities in data availability among countries partly stem from differing levels of willingness to disclose personal information.

Among those for whom we have information about their current residency, 46% lives abroad. Advanced countries, such as the USA and Japan, have most of their mathematical talents working in their own countries. More than three-quarters of medalists from Romania, Bulgaria, India, Belarus, and Armenia currently work abroad. Czech Republic and Slovakia, which recently became split countries, have differening fates for their medalists. Less than 20% of Czech medalists and more than 70% of Slovakian medalists live abroad. In general, the emigration of medalists depend on the income level of their home country. The rate of residing abroad is 62%, 49%, and 32% for low, medium, and high-income countries, respectively.

Table 7 The current residency of IMO medalists

Table 8 presents the countries hosting 1079 medalists who currently work abroad. There are ten countries that host ten or more foreign medalists. The medalists who live abroad are are primarily concentrated in the United States, with 54% of all medalists who live abroad residing there. In Table 6, we already listed seven countries that ten or more medalists obtained their foreign Ph.D. degrees. Notably, all seven of these countries are among the top eight countries that host the most migrant medalists. It is important to note that this may not be a casual relationship. Countries with the best universities also tend to offer high-value-added jobs that attract best talents.

Table 8 Destination countries for foreign residency

Figure 3 presents the ratio of foreign residency based on educational background. More than 70% of medalists who received their education abroad chose to stay abroad both at the UG and Ph.D. levels. In contrast, less than 20% of medalists who obtained their Ph.D. within their country chose to work abroad, while approximately 40% of medalists who completed their UG in their home country decided to work abroad. These figures indicate that education plays a significant role in emigration patterns. Countries that educate their own medalists tend to retain most of them in their home countries. Conversely, medalists with foreign educational backgrounds are more likely to work abroad.

Fig. 3

Percentage of medalists who live abroad conditional on their education

Non-academic jobs

Almost all academic careers leave a digital footprint on the internet. Therefore, it is highly likely that an IMO medalist who pursued an academic career can be easily located through an online search. However, there are exceptions to this rule. For example, an academic might change her name or transliterate it into Latin Alphabet in various ways. Out of the 2785 medalists, 1080 have pursued academic careers, so it is reasonable to assume that the remaining 1705 medalists are engaged in non-academic professions.

Figure 4 illustrates the distribution of fields for the 74% of non-academic medalists for whom we have information. More than half of these medalists pursued careers as software engineers, data scientists, or engaged in other computer related roles. Nearly a quarter of the non-academic medalists held occupations in the fields of economics, business and finance. Mathematics related jobs include roles such as olympiad trainers, high school mathematics teachers and mathematics content providers. Only 7% of non-academic medalists are employed in other science and engineering categories, such as medicine and mechanical engineering. Lastly, 4% of non-academic medalists are engaged in other occupations, such as music, law, and politics.

Fig. 4

Non-academic Jobs

There are only five companies that have attracted more than ten medalists. Not surprisingly, all five of these companies are either software or finance firms. Google stands out with the highest concentration of medalists, employing a total of 91 medalists, which is more than the total number of medalists employed by the remaining four firms. Microsoft, Meta, Citadel and D.E. Shaw & Co employ 27, 17, 15, and 13 medalists, respectively.

Academic jobs

We found that 1080 medalists pursued academic careers. We determined their academic fields using two different methods. First, we categorized academics based on the department in which they work. For instance, a researcher employed in a computer science department is assigned to the computer science field. Second, we determined the academic field based on the subjects of their publications in their Scopus profiles. We categorized all journals in Scopus into eight categories and assigned an academic to a specific category if the majority of her publications fell within that category. For example, if an academic had seven publications in mathematics, but fewer than seven publications in all other categories, then we categorized her as a mathematician.

Table 9 provides a comparison between the assignment by department fields and by publication subjects. It becomes evident that the results from these two methodologies do not always align. For example, only 70 out of 116 academics who are from computer science departments have the majority of their publications in computer science. However, both methodologies generally concur on the assignment of academics to the mathematics field. Of the 788 academics working in mathematics, 719 (91%) also have their majority of their publications in mathematics. Notably, 18 medalists who work in a mathematics department do not have any publications in Scopus-indexed journals, and 28 medalists who work in a mathematics departments predominantly publish in computer science journals.

Table 9 Academic fields by department type vs. publication subject

Next, we investigate the mathematics subfields in which 758 medalists, who have the majority of their publications in mathematics, specialize. There are 14 mathematics subfield categories in Scopus. We excluded “mathematics-miscellanous” category as publishing in that category does not indicate any specialization. We categorized medalists based on the subfield in which they have the majority of their publications. Table 10 lists nine mathematics subfields that has at least ten medalists. We see that the most crowded category is applied mathematics but more than a hundred medalists work in theoretical subfields such as algebra and number theory, and geometry and topology.

Table 10 Subfield that mathematician medalists specialize

We have previously observed that many medalists pursued academic careers, with a significant portion specializing in mathematics. Table 11 presents the ratio of academics who specialized in mathematics based on the income level of their home country. The highest ratio of medalists who became mathematics academics is found in the middle-income countries. More than a 20% of the medalists from low-income countries have managed to become mathematics academics. However, medalists from high-income countries enjoy advantages in all the bibliometric measures we examined. They have more publications and more citations than medalists from lower-income countries. The outcome may be influenced by the superior institutions in advanced countries. The disadvantage of being born in a low-income country cannot be overcome later in life when one becomes an academic. Nevertheless, the aspiration to become a mathematics academic is achieved by low-income medalists as well.

Table 11 Mathematics publication performance

Conclusion

We documented significant inter-country differences in the career paths of IMO medalists. For instance, there are 16 Romanian medalists who obtained undergraduate degrees from one of the top 5 global universities, while none of the Slovakian medalists graduated from these universities. There could be various reasons for this disparity. It is possible that domestic universities in Slovakia are better than those in Romania, leading Slovakian medalists to prefer staying in their country. Alternatively, it could be the case that high school education in Slovakia is of lower quality compared to Romania, resulting in top universities being less inclined to accept Slovakians due to this deficiency. It could also be the network of Romanians is more influential in securing placements at top global universities. Our analysis cannot distinguish among these potential reasons.

Countries seeking mathematical expertise for technological advancement should consider policy improvements to attract best talent from around the world to their universities. This is crucial because education serves as a key pathway for skilled migration. Since education is a strong driver of emigration, it is common for those educated in the United States to continue to work there (Kim et al., 2011; Stephan & Levin, 2001; Yuret, 2017).

Many medalists primarily chose the United States for their education. Despite the presence of many excellent institutions in various countries, they cannot attract international mathematical talents as much as a single institution from the United States. In most countries, there is usually one or two institutions that are preferred by medalists educated within that country. Therefore, it is essential to preserve the quality of these institutions if a country aims to retain and attract best talent to work within their borders.

Despite the limited opportunities for some medalists, there are optimistic lessons to be learned. Many medalists pursued their Ph.D. education abroad, and more than one-third of all the medalists succeeded to become academics. Around 70% of these academics chose to specialize in mathematics. From this, we can conclude that many medalists have the chance to specialize in their high school aspirations, irrespective of the income level of their country.

The career paths for medalists have consequences for the participants as well as their countries. The lack of opportunities for the best minds in a country poses a serious constraint for gifted children within that nation. If medalists cannot pursue the education that they aspire to, it is likely that less able students may also face limited prospects. In such cases, it becomes imperative to question the educational system of that country.

On the other hand, the inability to retain medalists within the country raises a serious alarm for the human capital stock of the nation. Human capital is typically measured by the number of college graduates, and medalists represent a sample of the most gifted students in that society. The migration of medalists may indicate a larger problem of losing the best minds in that society. The innovation and technological advancement of a nation often depend on the availability of these exceptional minds. In this context, the brain drain measured in our paper should serve as an alarming signal for the countries, guiding policies to retain their most talented individuals.

There are certain data limitations in our analysis. As we aim to compile a comprehensive dataset of all medalists, we encounter some missing data points. The more successful medalists are more likely to leave trace on the internet, introducing a bias towards a brighter career representation in our analysis. Additionally, missing data is culture-dependent, making our dataset more representative of cultures that are inclined to showcase their careers online.

Name confusions are inevitable. We have drawned the career paths of some medalists based solely on their name and high school graduation year. This method may inadvertently include a person with a bright academic career, sharing the same name and age as a medalist, if the actual medalist has no online presence. These biases tend to favor a more positive academic career presentation for the medalists in our analysis. Access to more information provided by national agencies could enable researchers to conduct analysis without such biases.

Another limitation of our analysis is that we are conducting a static analysis in a dynamic world. The world is constantly changing, and education opportunities, as well as global education opportunities have evolved through time. However, we have limited data points from some countries, making it challenging to perform a robust time-trend analysis. A future study that includes a more extensive representation of talented individuals from each country may better capture time-trends in career opportunities of gifted individuals.

We answered some questions in this paper but we have left many unanswered. The following questions should be answered in future research.

• Why countries of similar income and geographical location offer differing opportunities for their gifted children? While our paper highlighted intriguing binary comparisons, such as Romania and Slovakia, answering this question necessitates a closer examination of the educational systems and cultural aspirations of these countries.

• Are medalists content with their career decisions? Given the concentration of most medalists in the same fields of study and similar jobs, do they feel confined by their high school achievements? An interview with the Olympians should shed light on this question.

• How do the career paths of medalists differ from those who have not participated in the competition but have similar academic ability? To answer this question, one should select a comparison group of students, such as the top students from the best high schools in that country, and investigate their career opportunities. This approach allows us to discern whether the training and Olympiad participation influenced the career opportunities of the medalists.

• Why are some countries unable to achieve many medals in the IMO? For instance, there are advanced countries like Portugal that receive few medals. A closer examination of the educational systems of these countries is warranted to determine whether there are deficiencies nurturing the potential of their gifted individuals.

• Why are some countries unable to retain their medalists, and why do only a few countries manage to attract foreign medalists? As demonstrated in our study, the top five US institutions individually attract almost twice as many foreign educated medalists as all German institutions combined for Ph.D. education. A closer examination of admission and scholarship policies is necessary to unveil the reasons behind these discrepancies.

Appendix

See Table 12.

Table 12 Countries that have IMO medalists