You are on page 1of 27

International Comparisons in Higher Education

Funding

Domingo Docampo
ETSE Telecomunicación, Universidade de Vigo, Spain
email: ddocampo@uvigo.es

working paper: April 21, 2007

Chapter 1

Introduction

Education, by and large, is considered a public good, since it usually meets
the criteria of being both non-excludable and non-rivalrous1 . Consensus
around the ‘public good’ character of primary and secondary education is
overwhelming; as public goods, they must be made equally available to all:
they cannot therefore be provided for profit[Fis06]. Hence, in developed
countries (and some developing countries as well 2 ), state support reaches
every school so it can offer standard quality education (i.e. the best educa-
tion the system can provide regardless of the point of access) to the largest
possible share of the population3 . As such, these public goods can be sup-
ported by taxpayers and their delivery addresses both equity and efficiency,
since universal access and quality of service guarantee that nobody gets ex-
cluded, and there should be no easy and effective way of cheating to benefit
from a rivalrous choice.
Consensus breaks down at the point of tertiary education, particularly at
college education. According to standard economic theory, higher education
does indeed result in a large surplus of positive externalities, in the sense
that the educated person will benefit the society as a whole by producing
better goods, providing better services, and paying more taxes4 .
There is little or no discussion around the social value of Higher Ed-
ucation, given the positive externality that it spills over society at large.
However, from the point of view of the recipient of education, he or she
will end up being much better off, will enjoy more opportunities of career
building and enhancement, and will earn more money than remaining une-
ducated. It is then indisputable that higher education retains most of the
characteristics of a public good, but the private benefits accrued to students
are very apparent as well. Therefore it is argued that that kind of education

1

All States make provision for the students to pay their share. it should be related to the ability of the higher education system to accommodate all the aspirations of the youth. with New Zealand following suit) and late newcomers (East Asian countries).5% of their GDP in subsidizing higher education. 2 . a taboo. by looking at the subsidy one can deduce the value placed by governments on the positive externalities of higher ed- ucation. Nevertheless. to the desired extent. while a small number of them allow for a larger contribution from the students8 . that broad based questioning on the role of public funding in higher education is lacking in society at large [Fis05]. Equity is not per se attached to free education. Most OECD countries charge nominal fees for public higher education ser- vices. much less college education. regardless of their socio-economical background6 .8 and 1. students from disadvantaged socio–economic backgrounds to embrace tertiary education. as university officials and politicians alike recognize. with the OECD average situated at 1. The pertinent discussion would be how to evaluate the externality de- rived from each person’s education and compare it to the benefit accrued to the graduate.should be paid off at least in part by the students. and particularly in most European countries. countries spend in general between 0. Apart from the Scandinavian nations. unless universal access to higher education is in place. either up front or later on. Hence. However. Governments would have a way to internalize the positive externalities derived from education —that add social value to the private value accrued to its recipients— which is to evaluate the shift in the demand curve of higher education and introduce the subsidy that moves the equilibrium point close to the desired optimum. and so attached to ideology. It is unfortunate. the issue is so subject to partisan argument.3%. back in 1987 in Australia. that piling up evidence from sound scholarship over the past 20 years has not helped to open a public debate except in a few countries in the Anglo-American tradition (in the UK as recently as 2003. it should not be difficult then to share the costs accordingly. involving a significant transference of rents from the poor to the wealthy7 . once they graduate 5 . but true. By and large. there is a strong rationale for asserting that free higher education may be actually regressive. In continental Europe this is one of the non-debatable issues. government policies have not been that successful in encouraging. far too many developed countries show levels of access to higher education that lie below one third of the population at the applicable age. although the differences in quantity are very significant. Therefore.

Following [Bar04]. they include in the State subsidy the private value accrued to graduates. We will be assessing the role of governments in pro- moting higher education. there are some practical objections that proponents of free higher education must face. they will also have a say in the degree of private benefit which may be derived from higher education either to students or to institutions themselves[Fis06]. 1. persistent figures show how the debate on higher education financing is permanently unre- solved in those countries. In what follows we will summarize the discussion around the most difficult argument. Among European countries. but by under-financing universities. Those are the coun- tries that actually place a larger value on the returns of higher education. and very recently in the UK. the group of Scandinavian nations does assign a high social value to education in general and to higher education in partic- ular. that requires the inclusion of all citizens who either do not wish or are not able to participate in the benefits accrued by higher education. They encourage young people to pursue tertiary education. therefore. by stating the entitlement to higher education as a universal right. It is very difficult to question the status quo. But the debate took place in Australia 20 years ago. actually assign a lower social value to education. Other European countries. contradicting in practice what they state in principle. and make international comparisons on relevant indicators to find out whether actual policies are coherent with assumed po- litical stands. However.1 Is Higher Education a Truly Public Good? For some. We will first proceed with the discussion of the ‘public good’ character of tertiary education. The main and substantial argument goes as follows: Higher education should be wholly tax funded because graduates earn on average more (much 3 . The overall result is that in a number of countries the total effort in higher education exceeds largely the OECD average. excluding simple ideological positions as the one that asserts that higher education is a basic right and should therefore be free. In this paper we analyze the different approaches observed in tertiary education funding. for the same reasons that a similar argument could be employed on even more solid ground for food and shelter. and see then what international indicators tell us about the effect of the policies on the entry rate to universities. the idea of higher education as a public good is linked to the notion that without state intervention the market would fail to provide adequate provision for all citizens [Huf]. therefore.

which pay more taxes as well. What is interesting is that evidence shows how equity is not at stake in the majority of the countries that have taken that option10 . which collide with other priorities for public spending. And. A graduate pays then US$40. The same thing applies to education if it is believed that improving it is the only mean to face the challenge posed by globalization.more during a lifetime9 ) than non-graduates. education could become the first and foremost priority.000 less than a non-graduate with identical lifetime income. In the only country where higher education has never been free. It can be argued that there are countries in which income tax might be high enough as to render the first argument weak. at least in some countries. 3. showing that horizontal equity can at least be approached. the American public believes that a college education is a very valuable asset 4 . There are limits to taxation. there would then be a strong rationale for a heavily subsidized tertiary education sector. and that includes higher education. Some other countries have tried to improve the quality of higher education by seeking contribution either from students. Therefore. Were some or all of these conditions to coincide in a particular country. by virtue of the greater income taxes they return (again much more during a lifetime) grad- uates actually pay back for the education taken. 20% of which is deemed to pay for higher education. To this. Free higher education is horizontally inequitable. We shall see how the group of Scandinavian countries has consistently held public policies that make them good examples of a remarkably very successful public support of higher education. not least of political pressures. Barr offers some counter-arguments: 1.000 in tax. What about the unsuccessful ones?. By the way. 4. the consensus of public opinion around the importance of knowledge and the benefits accruing to society by research and development is beyond par- tisan debate in some countries. accordingly. 2. A person with a degree pays around an additional US$200. Besides. a very large percentage of the population does access higher edu- cation. or from graduates. The taxpayer gets ‘a good deal’ argument would also say that they (the taxpayers) should subsidize all development costs of successful firms. Income tax is paid by many more non-graduates than by graduates.

substantial public spending in higher education.2 Two approaches to higher education funding In [Bar04] the nature of higher education policies is characterized in terms of two ‘stylized’ models: In the ‘Anglo-American’ model. As far as diversity is concerned. we will be using data taken from the most recent international compilation offered by OECD on its Education at a Glance. The Anglo-American model should instead be attached to much lower taxes. It constitutes a well-known successful model of higher educa- tion.and wants as many people as possible to have access to it. The Scandinavian model would be one characterized by very high taxes. Hart and Teeter [HT03] conducted a national survey in 2003. since quality is much more easily associated with diversity than with uniformity. To confirm. and quality comparisons between them. What is most remarkable is that in spite of that divide. or not. one cannot deny the extraordi- nary diversity of the American higher education institutions. substantial private spending in higher ed- ucation and large enrolment figures as well. and encourages diversity. policy is based on the assumption that institutions are homogeneous. But it is not difficult to note that the phrasing of the models in Barr does not come free of implicit assumptions underlying the particular words chosen. and large enrolment figures. varied forms of provi- sions. But we could make an attempt to rephrase the two models in such a way that no unnecessary assumptions are made before analyzing the available evidence. there is neither a generation nor a partisan significant gap on this issue (right or privilege). and therefore treats them equally and regards all programs as equals. released in September 2006. a strong R&D commitment. 1. showing clearly how the American public opinion actually perceives both the public and private benefits of higher education11 . the characteristics posed. strong R&D commitment. arranged in a three-tier layer of research universities. capable of the best and of the least12 . 5 . In the ‘Scandina- vian model’. policy sees higher education as heterogeneous. and found that public opinion is divided about the status of higher education as a basic right or as something that has to be earned. teaching only colleges and commu- nity colleges.

without leaving out- side any relevant information. a number of continental European countries including the four largest among them. 6 . We would have liked to use the OECD entry rate indicator (percentage of youth in the corresponding age entering ter- tiary education) as an pointer to equity. we have chosen the Scandinavian countries. as it is obvious that the higher the entry rate the better the access opportunities to higher education. To reduce complexity in the treatment of the data. and countries in the Anglo-American track. based on the World Bank corresponding gross enrolment ratio: ac- tual number enrolled as a percentage of the number of youth in the official age group. The indicator of enrolment corresponds to the academic year 2003-04. The data in Education at a Glance has been compiled in 2006 but corre- sponds in general to 2003 or 2004. which render it practically useless for the present study. some Asian countries. We will instead anchor our analysis in the much more reliable Enrolment Indicator from the Economist [Eco07]. Unfor- tunately. at this moment the indicator presents severe inconsistencies in the distinction between type A and type B tertiary education students.

which measures funding to educational institutions by private sources. 1. Private expenditures in higher education correspond to OECD indi- cator B2. where. Public Expenditures correspond to OECD indicator B4. I5 Taxes on average worker I6 Enrolment I7 Gross domestic expenditure on R&D relative to GDP. I4 Total spending on higher education as a percentage of GDP. which mea- sures direct public expenditures on educational institutions plus public subsidies to households (which include subsidies for living costs) and other private entities as a percentage of GDP. I3 Private expenditure on higher education as a percentage of GDP. not only direct expenses on institutions are to be measured. 2.Chapter 2 International Comparisons 2. student help to cover living expenses should be accounted for as well.1 Data for the International Comparisons The data selected from the OECD tables concern the following indicators: I1 Public expenditure on education as a percentage of GDP.1b. We have selected in- dicator B4 for the reason that if equity is what is at stake. I2 Public expenditure on higher education as a percentage of GDP. 7 .

Enrolment corresponds to gross enrolment ratio: actual number en- rolled as a percentage of the number of youth in the official age group. We shall see how it is apparent from the plots that data confirms the existence of both the Scandinavian and the Anglo-American approach to higher education.8 Norway 7. NOR Norway.3 0. the following acronyms will be used: AUS Australia. CAN Canada. KOR South Korea. UK United Kingdom.6 3.4 1.1 shows the set of indicators for a number of OECD countries.9 81.6 16.2 Spain 4.6 41.8 0.1 2.5 1.1 0.0 1.1 2. JAP Japan.0 2.0 2.1 2. Indicators Country I1 I2 I3 I4 I5 I6 I7 Australia 4.0 1.2 38.0 0.0 Denmark 8.6 0.0 2.5 Italy 4.7 51.9 0.7 32.2 2. SWE Sweden.8 1. USA United States of America.6 85.0 1.0 2.8 88.3 58.5 67.6 Netherlands 5.6 58. NET Netherlands.0 3.0 1.2 20.2 0.5 2.0 United Kingdom 5.6 Finland 6.0 2.4 56.0 2.6 0.1: Main Indicators 8 .0 1.1 43.9 1.5 France 5.0 1.2 Korea 4.6 83.6 2. FRA France.2 1.7 77. DEN Denmark.4 26. We will then confirm the fact using appropriate statistical tools. We will now offer comparisons on relevant pair of variables that manifest the clustering of the countries.6 43. GER Germany.3 0.3 50.8 1.8 1.9 28. FIN Finland.1 Sweden 7.5 0.7 1. ITA Italy.0 1.0 4.3 48.2 0.2 Japan 3.7 Source: OECD Education at a Glance 2006 Table 2.6 2.6 51.4 47.7 0.1 1.0 2.8 New Zealand 6. 4. Income tax indicates tax on average worker (OECD economic indica- tor).1 0.1 29.4 36.0 2.7 57. Across the plots. Table 3.0 3.9 United States 5.3 1.2 0.5 2. SPA Spain.3 1.7 1.1 0.6 2.0 83.4 31.3 1.3 1.1 1.0 45.0 1.2 Germany 4.8 1.6 0.7 1.2 1.3 2.7 Canada 5. 3.0 62. NZE New Zealand.2 64.6 73.

00 Table 2.75 0.32 0. Public Spending in Higher Education Figure 2.45 0.1 shows the consistency of the public policies of funding for all the levels of education and for the particular segment of higher education.53 -0.2 clearly reflects the clustering of the four Scandinavian coun- tries around the highest positions in public spending and among the top in total spending in tertiary education.29 0.00 -0.90.00 -0.00 0. 0.25 0.40 0.52 0.39 0. and at some distance from the next three countries (USA. DEN NOR ! !! SWE FIN ! ! ! CAN !! NZE USA! ! NET ! !! GER! UK FRA SPA! !AUS ! ITA !! KOR JAP ! Figure 2.90 -0.69 0. Correlation coefficient is now 0.24 I2 1.34 I7 1. The Scandinavian countries score high in the two indicators at short distance from New Zealand. is very high in this case.08 I4 1.29 I3 1. Canada and France).2: Correlation Matrix Figure2.56 0. The divide is now clear. with Australia somehow leaving the group of lagers towards the Anglo-American model.00 -0.32 0.1: Public Spending in Education vs.41 0. Italy and Spain fall behind in these two indicators. Indicators Correlation I1 I2 I3 I4 I5 I6 I7 I1 1.00 0. It also shows how Korea and Japan.56. 9 .19 I6 1.34 I5 1. The correlation coefficient.00 0.48 0.

39). Total Spending in HE Now. USA KOR CAN DEN ( (SWE (( ((NOR ((( FIN NZE (( (( (AUS (((( NET (( FRA UK JAP SPA GER ITA Figure 2. but their record setting commitment to R&D places both countries slightly behind the regression line.clustered with the USA. we found only countries belonging to the two models. while Figure 2.32). USA CAN KOR DEN NZE NOR (((( (((SWE ( (( FIN AUS (((( (( (((( NET UK FRA GER JAP SPA ITA Figure 2.2: Total Spending vs. On top of the regression line (higher in total expenditures in tertiary education). which also combine high spending in tertiary education with strong commitment to R&D.4 points to a positive correlation between public expenditures in higher education and taxes on average workers (0.3: R&D expenditures vs. once total expenditures in higher education are taken into account (-0. with a correlation coefficient of 0.34.5 shows a negative correlation. Public Spending in Higher Education Figure 2. and Korea.3 shows relevant information on the relation between total spending in higher education and investments in R&D. 10 . Figure 2. We see the Scandinavian countries -but Norway that shows a worse record in R&D spending perhaps because of the recent dis- covery of natural resources. Finland and Sweden would also be there.

and the clustering is apparent. although the results for Denmark point to a deviation from the pattern difficult to explain. Norway. Taxes on average worker GER FRA SWE ITA NET FIN XXX DEN X SPAXX XXX NOR XXX UK XXCAN USA AUS XXX JAP XXX NZE KOR Figure 2. Correlation is significant (0. Australia and New Zealand. Figure 7 shows again the clustering of the Scandinavian countries. this time close to countries which made room for private spending in higher education.75 in absolute value) with low levels of taxes for the average worker. show the highest enrolment rates among all the countries. the Scandinavian countries. In both figures. 11 . Finland. Taxes on average worker Both figures point to the fact that private funding for higher education is highly correlated (0. combining large income tax levels with high spending (either public or total) in higher education. with Australia. Again.4: Public spending in HE vs. South Korea and Sweden over achieving and the USA. Figure 3. South Korea. GER FRA SWE ITA NET (( FIN ((DEN ((((( SPA ((( ( NOR ( (( ((( ( CAN UK USA AUS JAP NZE KOR Figure 2. USA. which one would expect to find on top of the regression line.6 speaks for itself. Denmark and remarkably Canada underachieving around the regression line. Finally. The same thing occurs to Canada. New Zealand. the cluster of the Nordic countries is very apparent.5: Total spending in HE vs.52).

South Korea and the USA). 0.6: Public spending in Education vs. Canada. Enrolment The divide between the countries in what we have called the Scandina- vian approach (Scandinavia) and the Anglo-American approach (Australia. Enrolment FIN KOR  SWE  USA NOR  NZE   AUS     DEN  UK  SPA  ITA  NET CAN  FRA GERJAP Figure 2. New Zealand. The correlation. and the rest of the coun- tries is clearly shown in this Figure.7: Total spending in HE vs. 12 . showing how increasing spending (either way public or private) has a great impact on enrolment and equity. FIN KOR USA SWE  NOR NZE    AUS      DEN SPA   UK   CANNET ITA FRA JAP GER Figure 2.82 is very significant.

and jointly explain a good deal (more than 76%) of the variance present in the data.74 0.81 0.5 38. The analysis was done using the main principal components —the ones showing an eigenvalue greater than one13 .86 0.5: PCA Regression Coefficients The indicators lie arranged as Figure 2.75 0.46 Table 2.88 0.8 -0. Component Eigenvalue % of var.55 Second Component 0. making the discriminator coming out of the two main principal components statistically sound.7 76.69 0. the higher (the closer to 1) the communality the greater the impact of the indicator in the selection of the axes or factors given by the principal component analysis.65 37.33 Table 2.89 0.89 0.47 0. 13 .69 38.5 2 2. Princ.5.2 Table 2. Indicators I1 I2 I3 I4 I5 I6 I7 First Component 0. The table of communalities14 is shown below.4: PCA Components As predicted. we will first perform a factor analysis on the set of indicators.86 0. The regression coefficients used to obtain the punctuation in both principal components are shown in Table 2. explained accumulated 1 2.96 0.3: PCA Communalities The table shows how all the indicators but I7 present a high communal- ity value.2 Factoring the two models In order to characterize the behavior of the countries on their approach or escape from what we have called the Scandinavian approach.97 0. the two principal components are very powerful.2.8 plots. Communalities I1 I2 I3 I4 I5 I6 I7 Final 0.

providing the signature for the Scandinavian approach.9 shows the perfect clustering of the Scandi- navian countries on the second quadrant.9: Principal Components Clustering of countries The plot of the two components for the countries included in the study is very apparent. Figure 2.8: Clustering of Indicators DEN SWE NOR FIN NZE USA CAN FRA NET AUS UK KOR GER SPA ITA JAP Figure 2. I3 I4 I6 I7 I1I2 I5 Figure 2. The UK does not conform yet with the model due to its weak figure for total expenses in higher education. The Anglo-American approach shows up on the first quadrant. because the variance within the Anglo-American approach is larger than the one shown by the Scandi- navian countries. 14 .10 shows now the results from both principal components. Figure 2. The clustering now is fuzzier.

10: Values of the first and second principal component: 38. respectively 15 .7% of the variance explained.5% an 37. KOR USA CAN AUS NZE JAP UK FIN SPA SWE NOR NET FRA DEN ITA GER SWE DEN FIN NOR USA NZE CAN KOR FRA AUS NET UK GER SPA ITA JAP Figure 2.

Table 2. I4.I2. We can see that the pair of axes selected by the statistical analysis does constitute a very good discriminator. can- didates to the Scandinavian model should get good positive scores in both components. Besides. We will treat both groups of coefficients separately. 16 . are committed to R&D and share large enrolment figures. points to a negative weight to public effort and high taxes. I6 and I7 correlate positively. The second set of coefficients. I3 and I5 point to the great divide in higher education funding. Each group of in- dicators should play the role of a unidimensional scale. It also accounts. and places the highest weight in private spending.6 shows the communalities. indicators I1. hence only the first eigenvalue should be greater than 1. while countries from the Anglo-American approach should get positive good scores in the first set of indicators (I1. KMO and Bartlett’s test from the principal component analysis for indicators I1. The first group of indicators is related to the common aspects of both Scandinavian and Anglo-American models. fund higher education properly (total spending). since both treat non-tertiary education as a public good.I6.2. and its relation to the level of taxes on the average worker.5 corresponding to the largest eigenvalue clearly signals to the combination of high public and total spending with large enrolment. The set of coefficients in Table 2. As seen in Table 2.I5). Indicators I2. eigen- values.I4. and I7 (let’s call them scale M1). slightly in the case of I7 though. discarding the data on private spending and taxes on the average worker.I7) and negative ones in the second set (I1.I6. to a lesser extent though. for total spending in Higher Education. corresponding to the second largest eigenvalue. both principal components from each scale (group of indicators) should be uncorrelated to guarantee a reasonable discriminant power. The remaining indicators measure the relation between public and private effort in higher education spending and among them and the level of taxing in the different countries. According to the characteristics assumed for the two approaches.I4. but shows some weakness in explaining accurately the reasons behind the clustering observed.

64 0.7 2 0.8 shows the coefficients of the indicators in the scales.83 0.15 53.7: Communalities.4 66.75 25 91. which confirms our initial hypothesis.56 0. Unidimensionality of the scales and appropriate values of both KMO and Bartlett tests15 to continue the factor analysis.006 Table 2. Coefficients Scales I1 I2 I3 I4 I5 I6 I7 M2 0. Communalities I6 I4 I7 I1 Final 0.60 12.99 66.87 0.006).7 53.75 0.04 Table 2. explained accumulated 1 1.6 3 0. explained accumulated Eigenvalues 1 2.75 0.63 0. freedom Significance 5.7 KMO BARTLETT Sphericity Test value Chi-Square deg.4 KMO BARTLETT Sphericity Test value Chi-Square deg.66 Table 2.7.I3.95 6 0. KMO and Bartlett Tets for Set 2 The preliminary analysis conforms to the characteristics predicted so far. Component Eigenvalue % of var.87 M1 0. Once the two principal components from both scales are computed.40 PC Eigenvalue % of var. and I5 (let’s call them scale M2) are shown in Table2.8: M1 and M2 PCA coefficients 17 .88 22 75.76 0. Communalities I3 I5 I2 Final 0.81 12.43 0.91 0. freedom Significance 5. their correlation is shown to be almost negligible (-0.4 2 0. Table 2.6: KMO and Bartlett Tets for Set 1 The result of the communalities of the principal component analysis for indicators I2.41 Princ.

we have two regions that summarize the discussion on the two models so far.11 to 2. although the UK has already initiated a departure from there in 2003.11: Principal Components M1 and M2 Scales 18 . the re- sults correspond to the Scandinavian model.2 clearly show the large land between countries within the two models and the rest of the nations analyzed. On the first quadrant. Figures 2. With the two uncorrelated scales. DEN SWE FIN NOR GER FRA NET ITA SPA UK CAN AUS NZE JAP USA KOR Figure 2. and countries within the fourth quadrant correspond to the Anglo-American approach (Korea included). It is within that large land where the major challenges rest.

lead to cope with the challenge of funding higher education while preserving equity. their data shows a remarkable gap between theoretical stands and actual policies. Unfortunately. leaving their higher education systems (and hence universities) in a weak position within the international context. the Scandinavian and the Anglo-American. 19 . which somehow deviates from the pattern in this particular indicator.13: Landing from the Future The analysis so far shows how both approaches. most of the continental European countries conform to neither approach.12: Landing from Scandinavia SWE USA FIN KOR DEN NOR NZE CAN AUS FRA UK NET JAP GER SPA ITA Figure 2. might be related to the accrued wealth from the oil reserves— confirming our initial exploratory analysis. We have also seen how R&D investments appear to be well linked to these two models —The case of Norway. DEN SWE FIN NOR GER FRA NET ITA SPA UK CAN AUS NZE JAP USA KOR Figure 2.

Scandinavian countries have been keeping close to the Scandina- vian model. Countries struggle to enlarge opportunities for accessing higher education but differ in how they would share the costs with students and graduates. Looking at the new rising stars in Asia.2. Many European countries make ambiguous claims about equality and learning opportunities. but student in- take has climbed rapidly over the last two decades and competing demands on government expenditure have squeezed spending on Education23 . since countries wishfully embraced expansion but few of them took the appro- priate policy measures24 . So far. In any case. unless firm resolutions are taken in setting up public spending in education as a true national priority or provisions are made to share the costs between students. and have shifted from talking about equity to actually delivering it22 . what we have called the Scandinavian model is solidly grounded. the countries that fall within the two models analyzed in the paper constitute successful education examples. Other countries theoretically assume the challenge of the Scandinavian model but fail to apply a coherent set of policies towards realizing its goals. graduates and taxpayers. although it might be the case that taxes are already too high20 and public expenditure will be more tightly scrutinized as the population age21 . OECD indica- tors show that the challenges of equity are being successfully addressed in the countries that have taken either the Scandinavian or the Anglo-American way17 . Two generations ago. and its educational standards were well below OECD average. an astonishing 97% of 25-to-34 year-old Koreans have completed secondary education —By far the highest rate among OECD countries19 — and its figures for tertiary education are impressive. Climbing the ladder of higher education investment is costly and lengthy. When the fruit is ripe students have proven not to be afraid of taking risks if the reward is dear and worth the effort. Today. They set ambitious goals to which others can aspire. partly reflecting funding agreements designed for a different era. it is 20 . By and large. Contributions from students and graduates are small or negligible.3 Final remarks Higher Education is on its way to become a universal aspiration16 . in a forthcoming paper we will show how the model is compatible with institutional diversity and quality. but evidence suggests that social background plays a larger role in determining a student’s performance in continental Europe than in Scandinavian or Anglo-American countries18 . Korea was counted among the poorest countries within OECD (its living standards were equivalent of today’s Afghanistan).

the rate of progression in China from primary to lower secondary school increased from 75% to over 98%. from lower to upper secondary education from 41% to 70%. The lat- est OECD’s report on Education at a Glance advices otherwise: accepting international bench-marking in educational performance is the basis for im- provement. The simplest example of a public good is an uncongested public highway. particularly in Europe. “the equity objective is not free education. and rose by 83% in Thailand and 51% in India”. according to recent indicators for many OECD countries. “[Education systems] must scale back the inherent class bias and sometimes catastrophically regressive way of funding existing edu- cational opportunities taxing the poor to subsidize educational opportunities for the rich in existing systems”. 3 Figures close to 100% in Elementary and Middle School and approaching the top in High School or equivalent. one cannot deny that there is a lot of truth in the observations that most gradu- ates earn significantly more than non-graduates and most students are from families that 21 . 7 Barbara Ischinger. John Kenneth Galbraith once said that given the choice of proving that changes are unnecessary. “Between 1990 and 2005. “It is only through such bench-marking that countries can understand relative strengths and weaknesses of their education system and identify best practices and ways forward”. “public universities. Notes 1 [Man98] page 220. 4 See [Man98] page 205. especially the flagship universities. and from senior high school to tertiary education from 27% to 76%. tables A1 and A2. If it is a toll highway then if it is not congested it is excludable but non-rival. and [Jon04] page 1. 2 [ST06] pp. but a system in which no bright person is denied a place because he or she comes from a disadvantaged background: a system in which the ability of the brightest students to study at the most intellectually demanding universities is unrelated to their socioeconomic background”. it becomes a natural resource. 5 See [Mar05] page 2. most people start working on the proof. If it is congested.undoubtedly likely that a change of attitudes will be necessary. provide not just collective public goods but high status private goods for their graduates: and that is the case even where university education is free of tuition charges” 6 [Bar04] page 266. the number of students attending university more than doubled in China and Malaysia. which benefits everyone”. Between 1995 and 2004 alone. “although some will disagree with part of the evidence. not seeking reasons why education systems should not or cannot be compared. “Excludability is the property of a good whereby a person can be prevented from using it while rivalness is the property of a good whereby a person’s use diminishes other people’s use. Public goods are neither excludable nor rival”. 26-27. to turn globalization from a threat to an opportunity. “Consumption of Education yields positive externalities be- cause a more educated population leads to a better government. and higher education funding will have to be part of it. it becomes both excludable and rival as any other common private good”. rival but non excludable. See for example [OEC06]. in [OEC06] page 18. but if it is congested.

“we estimate the median of the lifetime earnings distribution for male graduates to be around 600.400 institutions. It is both democratic and elite. 9 [OEC06] page 121.year-olds). “Attaining higher levels of education can be viewed as an eco- nomic investment in which there are costs paid by the individual (including reductions in earnings while receiving education) that typically result in higher earnings over the individual’s lifetime. For women. some institutions accept any applicant who is over the age of fifteen. with all countries showing a rate of return above 8%”. Forty-five percent of adults say that a college education is a right that should be available and affordable to all Americans. the final value is the one associated with the solution found. but it will never be confused with a system. 11 [HT03] page 8.000 more for women by comparison with non-graduate counterparts”. In this context. a caveat is opportune here.to 34. the initial value of a communality in a principal component analysis is 1. 15 Bartlett’s test of sphericity[Bar54] and the Kaise3-Meyer-Olkin (KMO) measure of sampling adequacy[Kai74] are generated by SPSS to help assess the factorability of the 22 .” 8 The majority of those countries seek a contribution from students but provide State loans to support further participation in college education. cites Australia. it was almost unnecessary since the un-rotated and rotated factors are almost identical. Party identification also creates a division on this issue.000 more for men and US$300. while older adults see it as a privilege (52% of adults age 65 and over). “American Higher Education consists of about 3. can produce private annual returns as high as 22. In addition. with more than half of Democrats (54%) believing that it is a right. After the iteration process to reach a solution takes place. USA and New Zealand along with the Nordic countries among other few nations as the ones offering an entry rate into university programs above 60% of young people. “Recent data show that average gross lifetime earnings for (Australian) university graduates are over US$450. 13 Although varimax rotation was used. while a handful of colleges rejects eight out of ten applicants”. It is indeed extremely difficult to prove any causal link between higher education and productivity as it is impossible to know whether a graduate is more productive simply by virtue of being a graduate. see [New04] page 6. they are divided on whether a college education is a right or a privilege. 1–2. the lifetime earnings advantage of the median graduate over the median non-graduate is around 800. Nevertheless. during their lifetime.may be regarded as more advantaged than others”.0 00 US$ higher than the equivalent for male non-graduates. “Certainly there is evidence of a positive correlation between higher education expenditure and per–capita income growth but no clear evi- dence of a causal relationship between the two.6%. page 270. 14 Communalities actually measure the proportion of each variable’s (indicator in our case) variance that is explained by the principal components. any analysis relies on formal qualifications and wages as measures: this is problematic as qualifications” 10 Education at a Glance.000 US$”. and about half half (52%) of Republicans seeing it as a privilege” 12 [HM05] pp. See also [DES05] page 8. when undertaken as part of initial education. and 41% say that a college education is a privilege that must be earned. “while most Americans agree that access to college should be widened. See also [DFG+ 05] page 9. Also [Bar04] “Taxpayers subsidies are regressive and free higher education has done bad on access. converted to US $. Orig- inal figures in British Pounds and AUS$. the investment to obtain a university level degree. There is a generation gap on this issue (younger adults see a college education as a right (53% of 18. By definition.

5. 21 See Lundsgaard.6 is suggested for a good factor analysis[TF96]. Study now. 16 See Adrian Wooldridge. The brain business. page 31. The minimum value to proceed further is 0. 23 . 19 ibid. September 8th. Survey on higher education: The Economist. and Turner. The attitudes of European societies towards this. Education at a Glance 2006.5). page 26 20 See [Coo04] page 155. page 18. “Continental European countries are close to the Anglo-American limits of taxation. 22 See Adrian Wooldridge. showing a readiness to make the changes necessary changes to create the opportunities. OECD Observer. Many governments have tried to square the circle through tighter management. forcing Europe to compete on basic products as well as very sophisticated ones with a high knowledge content. J. “If more and more governments are embracing massification. 24 See Barbara Ischinger.data. will determine whether global competition becomes more of an opportunity than a threat for them”.. “The conservative argument (it is foolish to waste higher educa- tion on people who would rather study ‘Seinfeld’ than Socrates) falls at the first hurdle: Anglo-Americanity. The KMO index takes values between 0 and 1. with the dramatic rise of the new Asian dragons. 17 See Barbara Ischinger. few of them are willing to draw the appropriate conclusion from their enthusiasm: that they should either provide the requisite funds (as the Scandinavian countries do) or allow universities to charge realistic fees. but management cannot make up for lack of resources”. To consider the factor analysis appropriate the sphericity test should be significant (p < 0. elephants and tigers presents a challenge. 23 See [Mar06]. pay later.. 2005. Higher education is rapidly going the way of secondary education: it is becoming a universal aspiration”. page 37. D. page 18. “The situation facing Europe today. and 0. 18 See [ST06]. March 2004. The closer to 1 the better. around 50% of GDP”. Education at a Glance 2006.

Bartlett. 25(2):157–72.pdf.Bibliography [Bar54] M. 2004. [DES05] DEST. 2007 edition. Higher education funding.dit. Journal of the Royal Statistical Society. 2005. [Bar04] N. chapter A Glimpse of 2020. Barr.uk).au/reforms.gov. [Eco07] The Economist. Goodman. Fisher. [Fis05] S. webpage: http://level3.ie/html/issue3/fisher/fisher. Profile Books Ltd.ifs. 1954. Fisher. 2005.htm. Pocket World in Figures. Palgrave. The Global Competiveness Report 2004-05: World Economic Forum. Oxford Review of Economic Policy. Commentary 98. 2007. of Lifelong Education. Dearden.. Technical report. Higher education funding policy: Who wins and who loses? A compre- hensive guide to the current debate. J. . Our Universities: Backing Australian Future. (16 (Series B)):296–98. Fitzsimons. [Fis06] S. 2004.. [DFG+ 05] L. webpage: www. [Coo04] R.org. Cooper. 2005. A note on the multiplying factors for various chi- square approximations. 24 . 2006. A. and G Kaplan.S. Dublin Institute of Tech- nolgy.backingaustraliasfuture. Australian Government. The Insti- tute for Fiscal Studies (www. MacMil- lan. 28(2):265–83. E. Is there a need to debate the role of higher education and the public good? Technical report.N. Does the Celtic Tiger Society need to debate the role of higher education and the public good? Int.

Merrow.hefce. 2004. Colin Bell Memorial Lecture. Rethinking public/private in higher education for the global era. Jongbloed. Palgrave MacMillan. Philadelphia. In Fulbright brainstorms 2004 –new trends in higher education. [OEC06] OECD.education. [Kai74] H.au/ centres/mcrie/docs/otherrecentpapers/ ashe-2005-colloquium-p-and-p-pres-version. 1998. Funding higher education: options. dragons. affordability and access: Americans speak on higher education. Declining by Degrees: Higher Education at Risk. Psychometrika. (39):31–36. [Jon04] B. The Dryden Press. USA.eu/pdf/challenge2006.PDF.edu/ uploads/document-library/789/10C–QUALITY -AFFORDABILITY-AND-ACCESS—ETS. [HT03] P.epc.pdf. 1974. Quality.ac. Martens.monash. Higher Education in Europe. Newby. Challenge Europe. elephants and tigers: adjusting to the new global reality. Schleicher and K. webpage: www. [Mar05] S.[HM05] R. 2005.M.0/pdf/15 benjongbloed. dragons. [Mar06] H.doc. pages 37–41. 2005. chapter I-4: Living in the Past?.nodak. Mankiw. Tremblay. University of Bradford. (3):339–348.pt/brainstorms/release1. chapter 25 . Doing widening participation social in- equality and access to higher education. Higher education as a public good: means and forms of provision.edu. [New04] H. Hart and R. 2006. [Man98] N. 2006.ccla. Kaiser.pdf. [ST06] A.G. Herhs and J.uk/news/events/2004/bell/bell. Challenge Europe.pdf. webpage: www. 2003. Marginson. trade -offs and dilemmas. ele- phants and tigers: adjusting to the new global reality. ASHE 2005. Teeter. Hufner. webpage: www. Principles of Economics. European Policy Center. In Higher Education Colloquium.ndus. webpage: www. Educa- tional Testing Service. webpage: www. An index of factorial simplicity. Technical report.D. 2004. [Huf] K. Education at a glance.

pages 24–36. 26 . I-3: Education and the knowledge economy in Europe and Asia. [TF96] B. New York: HarperCollins. Tabachnik and L.S. Fidell.G. European Policy Center. 2006. Using Multivariate Statistics (3rd edition). 1996.