Low Earnings For High Education Greek Students Face Weak Performance Incentives Wasilios Hariskos, Fabian Kleine, Manfred Königstein & Konstantinos Papadopoulos 1 Version: 19.7.2012 Abstract: The current economic crisis in Greece influences expectations of young people regarding their own long-term opportunities. To investigate this issue we ran a questionnaire study that compares expectations of Greek and German students regarding income development and the difficulty of finding an adequate job. Greek respondents expect only a small earnings differential of highly educated compared to low educated employees. Also, they expect great difficulties in finding an adequate job after completion of their studies. Thus, the incentives to attain a university degree are relatively low. Hariskos, Kleine and Königstein are affiliated with University of Erfurt, Applied Microeconomics, Nordhäuser Straße 63, D-65611 Erfurt. Königstein serves as corresponding author: manfred.koenigstein@uni-erfurt.de. Papadopoulos is affiliated with Aristotle University of Thessaloniki.
1 Introduction What do Greek students think about their economic future in the face of the current economic crisis? How large are the income losses they expect? What do they think are their economic benefits of having a university degree? How long do they expect to search for a job that matches their qualification? And, how large is their willingness to look for a job abroad? These questions point to serious long-term effects of the current crisis on the Greek economy. Different age groups and income categories are affected differently. The rise of unemployment has affected the young and middle aged the most, while wage and pension cuts as well as multiple new taxation laws have led many Greek households into deficit. Greek university students constitute a special social group. They are the future educated workforce of their country and unambiguously form an essential value-adding part of the Greek society. Their expectations about the state of the Greek economy and the development of labor market conditions may indicate the imminent braindrain. On November 24th 2011 we ran a questionnaire study to address the above questions. Specifically, we collected individual data on expectations regarding the development of net incomes, duration of job search, and international labor mobility from students at the Aristotle University of Thessaloniki, Greece and students at the University of Erfurt, Germany. We are primarily interest in the data from Greek students. The data of German students serve as an interesting benchmark. Germany is less hurt by the euro zone debt crisis, so student expectations in Germany are formed in a more stable environment. Moreover the data from German students can serve to control for systematic errors in the formation of expectations. The main results of our study can be summarized as follows: Firstly, Greek students expect rather substantial income losses for both, persons with average education as well as the highly educated. Moreover, the absolute and relative benefits of higher education are viewed as rather small. Secondly, Greek students express a much stronger willingness to look for a job abroad than German students. This might be due to the fact that Greek students expect a much longer time of search for a job that matches qualification. The paper proceeds as follows: The next section outlines the situation in Greece around the time of our data collection. Thereafter (section 3) we explain the construction of the questionnaire including our ideas underlying the questions. Section 4 presents data analyses and results. Section 5 concludes. 2 The economic and political environment in Greece in November 2011 In 2011 Greece faced a deep recession for the fourth consecutive year with real GDP growth expected at -5.5% by the end of the year and another decline of -2.8% estimated for 2012. Unemployment was about 16.5% out of the total workforce of 5 million and reaching 42% among young people of age 15-24. Government was expected to lay off 200,000 employees in the public sector until 2015, implementing a special labor reserve system that would push unemployment even higher. A number of new taxes and tax rate increases were implemented, including a solidarity tax and a real estate tax. At the same time wages had fallen by about 25-30% on average with respect to their 2009 level. Further austerity measures included closure or merger of public entities, and deep cuts in salaries and special benefits of public-sector employees. According to the Greek budget draft published in November 2011, central government debt was 383 billion (ESA 95) representing 174% of a GDP of 220 billion in 2011. For 2012 these numbers were
expect to rise to 400 billion (186% of a GDP of 215 billion) despite efforts for fiscal adjustment and structural reforms agreed between the Greek government and EC/ECB/IMF in 2010. The european summit in October 2011 concluded that a deeper private sector involvement was needed in order to render Greek debt sustainable aiming at a 120% by 2020. After the summit the announcement of Prime Minister Papandreou to have a nationwide referendum on the governmental policy triggered a series of reactions at the european and the national level and finally lead to Papandreou s resignation. Ex-Vice ECB President Lucas Papademos took over as new Prime Minister, backed by a tripartite coalition between the PASOK socialists, the New Democracy liberals, and right-wing LAOS. By end of November 2011 it was uncertain whether the Eurogroup members would agree to pay the 6th tranche of the May 2010 loan (8 billion Euros). Money reserves were estimated to run out around mid of December. Thus, for the Greek public the threat of bankruptcy was evident. 3 Construction of the Questionnaire Part 1 of the questionnaire (questions 1.1 to 1.6) asked the students for their perception of the development of income and taxes (including social security contributions), for their perceived difficulties in finding a job in their home country and for their willingness to look for a job abroad. A key statistic of individual well-being is the yearly net income a person receives. Therefore we aimed at getting subjective expectations of students with respect to this figure. Simply asking students for their expectations on the development of net income seemed problematic since the participants might confuse gross income and net income. Therefore we decided to provide a benchmark case for gross income and taxation (income tax and employee contribution to social security), first, and to ask for the development of gross income and tax rate, second. We can use these responses to calculate expectations on net income. Specifically, we informed that EUROSTAT reports for the year 2009 that the average gross earnings per year of single employees without children in Greece are 18,541 EUR (40,929 EUR in Germany). Furthermore we informed that EUROSTAT reports a tax rate (including employee contributions to social security) of 20.82 % (41.32 % in Germany). Based on these information students were asked to provide, e.g., an estimate of the development of these statistics for the years 2011 and 2013. In the following we refer to the students estimates as expectations. By these means, we think, we avoided the problem of participants confusing gross and net income. Furthermore, we think it is advantageous that we provided a common baseline for forming expectations. And, of course, gross income and tax rate can be used to calculate expectations on net income. Question 1.1 asks for the development of income and taxes of an average (benchmark) person in 2011 and 2013. These responses tell us the perceived development for the two years before our study and the next two years to come. We decided to ask for their expectations of 2011 and 2013 to collect both, their assessment of the changes that have happened in the past and those that may come in the future. Question 1.2 asks for the development of income and taxes of employees holding a university degree. A comparison of 1.2 with 1.1 informs about the perceived relative income development of higher educated persons compared to average persons. After completing their studies the students will belong to the highly educated part of the workforce. So expectations on this cohort may indicate even stronger the development and future economic chances seen by students. Question 1.3 asks the respondents for an assessment of their own expected income situation. They have to enter the expected year of entering a job entry after graduation and the expected gross income in the first year after job entry as well as the tax rate. The question asks for income and tax
conditional on job entry to avoid problems caused by expected times of unemployment if we had simply asked for income in the first year after graduation. Question 1.4 measures the perceived difficulty of finding an adequate job within the home country. We asked to consider a job that fits qualification since being employed at all is not the only important aspect here. Well-being and job satisfaction depend also on the job matching personal qualification. Finally, questions 1.5 and 1.6 ask for student s own willingness to look for a job abroad and for the student s assessment of the average willingness of other students to do so. These questions can be used to measure the brain drain one might expect for the future, i.e., the loss of highly educated workforce due to migration. Questions 2.1 to 2.5 are used to collect information on age, gender, nationality, number of semesters studied, and the field of study which will be used for partial analyses when adequate. Finally, the third part of the questionnaire (questions 3.1 to 3.10) implements the lottery experiment of Holt and Laury (2002). It can be used to extract individual attitudes toward risk. Individual risk parameters may be used in addition to other personal characteristics (gender, etc.) as explanatory variables in regression analyses. Since the future economic development is uncertain we thought that risk attitudes might play a role. 4 Results 4.1 Dataset The data comprise the questionnaire responses of 234 students of Aristotle University of Thessaloniki (sample Thessaloniki or Greece) and 87 students of the University of Erfurt (sample Erfurt or Germany). The Greek data were collected on Nov. 23 rd 2011 and the German data were collected on Nov. 24 th. Table 1 displays statistics on sample characteristics (gender, nationality, field of study, age and number of semesters studied). Table 1: Personal Characteristics Greece Germany Sample Size 234 87 Male 52 % 52 % Native 95 % 99 % Age 19 89 % 26 % Semester = 1 92 % 91 % Economics 96 % 37 % Most respondents study economics and are in their first semester. About half are female. Greek students are somewhat younger than German students reflecting the different school systems leading to a lower university entry age in Greece. Median age is 18 in Thessaloniki and 20 in Erfurt. While almost all students in Thessaloniki (97.0 %) study economics the sample is more mixed in Erfurt. Here only 36.8 % of respondents study economics as major or minor. However, many respondents reported Staatswissenschaften which is an interdisciplinary study program comprising three fields (economics, law and social sciences) within the first year of study. Together this makes up to 89.8 % of Erfurt students studying at least some economics. In the following unless stated otherwise all data analyses will be based on native students; i.e., we use only Greek respondents of the Greece sample and German respondents of the Germany sample. This comprises 96.3% of all respondents (95.3% of the Greece sample and 98.9% of the Germany sample).
4.2 Expectations on the Development of Net Income Table 2 summarizes expectations on the development of net income. It shows median responses for both samples, Greece and Germany, separately. 1 Table 2: Net Income for a Single Household Year Greece Germany EUROSTAT 2009 14,681 24,017 Expectations 2011 11,790 24,063 2013 10,600 24,009 2009 15,200 30,000 Expectations 2011 12,226 30,117 (University Degree) 2013 10,640 29,965 EUROSTAT Poverty Threshold 2010 7,178 11,278 Accordingly in Greece net income of single households without children is expected to decline by about 2,900 EUR from 2009 to 2011 and by about 1200 EUR from 2011 to 2013. In total this is reduction of 27.8 % (from 14,681 EUR to 10,600 EUR). On the contrary, German students expect net income to stay about constant. The differences between expectations of Greek students and German students are statistically significant at the 0.1 % level according to Mann-Whithney-U-tests (p < 0.001). 2 The expectations on net income of persons with university degree show even stronger differences between Greek and German students. Again the students expect declining income in Greece and stable income in Germany. What seems even more important is the expected earnings differential of highly educated persons compared to average persons. In Greece the difference is about 3.5% in 2009 (15,200 EUR compared to 14,681 EUR), declining to 0.4 % in 2013. In Germany the difference is stable at about 25 %. A Wilcoxon-Matched-Pairs-Signed-Ranks-test shows significantly higher income for higher educated persons for both samples (TH: p = 0.028, EF: p < 0.001). However, the difference is economically substantial only for Germany. We conclude that German students expect substantial benefits of education whereas Greek students don t, at least if they plan to find a job in Greece. The data are illustrated by Figure 1 showing the net incomes 2009, 2011, 2013 relative to the national poverty threshold 2010 for singles provided by EUROSTAT. The poverty threshold is defined as 60 % of median equivalized income after social transfers. Households with incomes below this cutoff are regarded as poor. The poverty threshold 2010 for singles is 7,178 EUR for Greece and 11,278 EUR for Germany (see Table 2). 1 We report median responses instead of mean responses for robustness against outliers. 2 Unless stated otherwise all statistical tests reported are two-tailed tests.
Figure 1: Net Income Relative to National Poverty Threshold 2010 German students expect average income to stay well above 200 % of the poverty level (above 250 % for the highly educated). For Greek students the values are above 200 % of the poverty level for 2009 and fall below 150 % for 2013; i.e., the earnings differential compared to the 2010 poverty level is expected to fall by more than 50 percentage points! A methodological point to mention is the following: We choose to compare expected incomes to the 2010 national poverty levels even though EUROSTAT calculates poverty thresholds on a yearly basis and the thresholds change over time. If Greek incomes decline, this will reduce the Greek poverty threshold, since the poverty concept used by EUROSTAT measures relative poverty instead of absolute poverty. This implies that a decline in incomes does not need to influence the poverty rate (fraction of the population considered as poor). It is even possible that the poverty rate declines. This happens if median income decreases more than income in low income classes. However, we choose a fixed poverty threshold to illustrate how close income of the relatively well-off population in Greece is expected to move toward a level that was considered as poverty level only some years before. The results on the development of net income can be summarized as follows: Greek students expect rather substantial income losses for both, persons with average education as well as the highly educated. The absolute and relative benefits of higher education are viewed as rather small. Average net income is expected to fall from more than 200 % of the 2010 national poverty level to less than 150 % of that level. On the contrary German students expect a stable income development that stays above 250 % of the 2010 national poverty level for the highly educated. 4.3 Job Search and Mobility Table 3 shows statistics on job search and mobility for both samples. Table 3: Job Search and Mobility Greece Germany Year of Job Entry 2017 2016 Net Income at Job Entry 10,500 23,013 Search Adequate Job 2,0 years 0.5 years I Leave Country 44.6 % (2.0 %) 37.4 % (2.5 %) Others Leave Country 49.2 % (1.2 %) 33.9 % (1.7 %)
The expected year of job entry (median) is 2017 in Thessaloniki and 2016 in Erfurt. Greek students expect to earn 10,500 EUR in their first work year which is about the average income expected for 2013. German students expect to earn substantially more (23,016 EUR) than Greek students, which is in line with the analysis on expected income development presented in the last section. But this value is well below the expected income level of highly educated in 2013. Thus, students at Erfurt seem to correctly anticipate that in Germany earnings strongly rise with seniority. An alternative explanation could be that students expect average gross income to decrease or taxes (including social security contributions) to increase rather substantially between 2014 and 2016. But this explanation seems implausible given the expected development till 2013 reported in the last section. Table A.1 in the Appendix reveals that the tax rate at the time of job entry is expected at about the same level as before. Rather it is gross income at job entry which is expected to be smaller than the average level of highly educated persons in 2013. Given the stability of expectations for the time 2009 to 2013 we rule out the possibility that students at Erfurt expect a strong long-run economic decline and favor the seniority-effect as explanation. For the Thessaloniki data a seniority-effect is hardly visible, which is not surprising given the observable pay compression: highly educated persons are expected to earn about the same as average persons. Furthermore, Table 3 reports a rather long time necessary to find a job that matches qualification (median of 2 years in Thessaloniki). This indicates that many students expect their first job not to meet their qualification. This is different at Erfurt where students expect to find a job that matches qualification within 6 months. In the lower part Table 3 reports expectations on expected migration. It displays the respondents own willingness to look for a job abroad expressed as probability (see I leave country in Table 3 and question 1.5 of the questionnaire). The mean probability is 44.6 % for Thessaloniki and 37.4 % for Erfurt. The difference is even more pronounced for the respondents assessment of the average willingness to look for a job abroad within the student cohort (see Others leave country in Table 3 and question 1.6 of the questionnaire). Here the mean probabilities are 49.2 % for Thessaloniki and 33.9 % for Erfurt. Table 3 reports relatively small standard errors of means indicating that the differences between Thessaloniki and Erfurt are statistically significant. This is confirmed by Mann-Whitney-U-tests rejecting the Null-Hypothesis ( No difference between Thessaloniki and Erfurt ) at significance levels 5% (p = 0.028) for I leave country and 0.1% (p < 0.001) for Others leave country. Since the standard errors are smaller for Others leave county than for I leave country the former might be viewed as more reliable information. Accordingly expected job migration is large in both, Thessaloniki and Erfurt, showing a high mobility of young educated persons in both countries. However, that the probability is 15.3 percentage points larger for Thessaloniki indicates a substantial brain drain for Greece. Expressed differently the mobility measure for Thessaloniki is 1.45 times the value of the mobility measure for Erfurt. The results on job search and migration can be summarized as follows: Greek students express a much stronger willingness to look for a job abroad than German students. For the TH sample the average expected probability of leaving the country is 1.45 times the corresponding value of the EF sample and reaches almost 50 percentage points. The higher mobility of Greek students might be due to the fact that they expect a much longer time of search for a job that matches qualification.
5 Concluding Remarks The current economic crisis in Greece has a strong negative effect on young people s expectations regarding their own economic opportunities in the future. While this result was to be expected it is interesting to look at effect sizes and compare the Greek responses to the German control group to get a relative assessment. Greek students expect substantial losses in absolute and relative terms whereas German students expect stable conditions. Monetary incentives to get a University degree are particularly flawed. Greek students expect to earn about the same as people with lower education while German students expect a substantially positive and persisting earnings differential. In particular, earnings expectations are dramatically different in relation to national poverty levels. In Germany earnings of the highly educated are expected at more than 250% of the 2010 poverty level, whereas these earnings are expected to fall to about 150% in Greece. The willingness of Greek students to leave the country for a job abroad is about 1.45 times as high as the willingness of German students. The probability of leaving is around 50% in the Greek cohort which implies a major brain drain. Our study highlights the economic decline that Greeks are facing currently and for the years to come. By now (July 2012) the situation has actually deteriorated compared to the time of the study (November 2011). It seems that the students expectations were realistic. Unfortunately, our data don t allow us to draw a more optimistic picture.
Appendix Table A.1: Median Expectations on Gross Income Thess Data Median Erfurt Data Median EUROSTAT gross income 2009, single 18,541 40,929 Q1.1 income 2011 15,500 41,311 Q1.1 income 2013 14,000 41,500 Q1.2 income 2009 19,000 55,000 Q1.2 income 2011 16,000 55,250 Q1.2 income 2013 14,400 55,200 Q1.3 income job entry 13,200 40,750 Table A.2: Median Expectations on Tax Rates (Including Empl. Contrib. to Social Security) Thess Data Median Erfurt Data Median EUROSTAT tax rate 2009, single 20.82 41.32 Q1.1 tax rate 2011 23.00 42.00 Q1.1 tax rate 2013 24.11 42.52 Q1.2 tax rate 2009 20.30 43.00 Q1.2 tax rate 2011 22.45 44.00 Q1.2 tax rate 2013 23.00 45.03 Q1.3 tax rate job entry 22.00 42.50