3. The operating grant for Flemish higher education institutions

In recent decades, Flemish higher education policy has been guided by the principles that higher education should be widely accessible for the population, entail low costs for students seeking initial qualifications at each level of the system, and be flexible enough to allow students to study at different paces, depending on their personal circumstances (Flemish Government, 2015[1]).

In common with the systems in Austria or the Netherlands, higher education in the Flemish Community is largely “open access”. As a rule, admission to initial programmes (at associate degree [graduaat] or bachelor’s level) requires students to have successfully completed upper secondary education or to have an equivalent level of competencies formally recognised, but it is not dependent on obtaining specific grades in school examinations or passing entrance examinations at system or institutional levels. Exceptions to this rule exist for medicine, dentistry and programmes in the performing and visual arts in the Schools of Arts. In these cases, entrance examinations, with pre-established minimum scores, are used to select students for admission. In all cases, to access master’s programmes, students require an academic bachelor’s degree in a relevant field or to have successfully passed a relevant bridging or preparatory programme.

The Government is committed to contributing financially to the cost of an initial bachelor’s degree and master’s degree for every eligible individual who meets the relevant requirements for admission to higher education (Flemish Government, 2015[1]). The Government also provides institutional funding for students taking initial associate degrees, which, as discussed, have recently been integrated into the higher education system. In principle, the Government’s funding commitment covers eligible individuals, of any age, who do not already have an associate, bachelor’s or master’s degree and can meet the requirements to enter or re-enter the system. The comparatively low tuition fees in the Flemish Community at undergraduate and post-graduate level (see Chapter 5) mean that higher education institutions are dependent on public funding to finance almost the entire cost of initial education programmes for eligible students.

All programmes in the Flemish higher education system are structured into courses, each worth a specific number of credit units (studiepunten). Students have considerable flexibility in the number of credits for which they enrol each year. The vast majority of students enrol in a “diploma contract”, with the explicit intention of obtaining a degree. However, it is also possible to enrol in higher education on a “credit contract”, under which students take only a specific number of course credits, without the intention of acquiring a full qualification. Students enrolled in a “diploma contract” must take specified modules that will permit them to obtain a qualification in their chosen field, and institutions provide guidance on recommended annual credit loads. However, within limits, the credit-based system makes it possible for students to combine programme learning components more flexibly than in systems with rigidly defined study programmes, and to retake modules or switch programmes during their higher education career with relative ease.

To support this flexible system of higher learning, the Flemish Government has designed its institutional and student-funding models with an explicit link between payments and the number of credit units for which students are enrolled or that they successfully obtain. As part of this system, all eligible individuals receive a “learning credit” account with credits that they can “spend” on higher education (see Box 3.1).

The objective of the learning credit system, when it was introduced as part of a wider reform of the funding model in 2008, was to maintain and strengthen flexible conditions for accessing higher education, while also placing a limit on the length of time that an individual student could remain in higher education at taxpayers’ expense (Flemish Government, 2015[1]). The flexibility to enrol for different numbers of credits can facilitate part-time study for those in work, with caring responsibilities or with disabilities, for example. The 60-credit bonus that students receive when they successfully pass their first 60 credits in a degree programme was designed to reward success, but also allow students who fail a proportion of courses in their initial period in higher education to maintain sufficient credits to continue their studies (Flemish Government, 2015[1]). A student who fails 60 credits before completing a bachelor’s programme successfully, for example, will finish their bachelor’s with a learning credit balance of 140, which is enough to allow them to complete a full master’s degree with margin for failure. Ultimately, however, students who repeatedly fail to pass credits will use up their learning credit, creating – in principle – an incentive for careful study choice and commitment to studying.

In practice, the system of learning credit has been criticised for providing limited incentives for students to progress swiftly through higher education. In particular, there is no maximum limit on the time students can take to pass their first 60 credits, and thus qualify for the 60-credit bonus, and the overall level of credit allows students to spend several years in more than the theoretical programme duration (three years for bachelor’s and two years for master’s) to acquire their qualifications. As discussed later in this chapter, the weak incentives for intensive study in the early stages of students’ higher education careers, as well as slowing progression, may be a factor in the comparatively high rates of drop-out from Flemish higher education, noted in the previous chapter (Werkgroep "Studievoortgangbewaking", 2014[4]).

The Flemish Community uses a formula funding model, introduced from the academic year 2008-9, to allocate the available budget envelope for public core funding to higher education institutions to individual universities and university colleges. The formula is used to calculate the size of the grant each institution receives each year, with payments to institutions made as a lump sum, which they are then free to re-allocate internally. As explained in more detail below, the size of the lump sum grant for each institution is driven to a large extent by the number of credits for which students enrol in each institution and the number of credits that they successfully pass, using data from previous years as a calculation basis. As such, there is a strong relationship between (past) student numbers and the funding institutions receive.

In student-driven funding systems, such as that in the Flemish Community, the concept of “fundable student” is important. In the Flemish system, students must fulfil three main criteria to be considered “fundable” and thus for their data to be included in the formula calculation and for their host institution to receive public funding for their education:

  1. 1. They must be enrolled in a programme or course that is eligible for funding. This includes accredited associate degree, bachelor’s and master’s programmes, bridging and preparatory programmes, as well as individual courses that are part of these programmes, but in which students can enrol outside a formal degree programme on a “credit contract”. The formula funding system provides a full rate of funding per credit for students who enrol for the first time in a particular programme type (e.g. their first associate or bachelor’s degree). The funding model also provides funding at half the full rate for students who take specialisation programmes in university colleges (bachelor-na-bachelor), although no institutional funding is provided for post-master’s specialisation programmes in universities (master-na-master). In the latter cases, students pay fees that are higher than the statutory level mandated for fully funded programmes.

  2. 2. Students must hold the nationality of a country of the European Economic Area (EEA) or otherwise fall into limited number of other categories, primarily related to refugee or migrant status. Flemish legislation allows higher education institutions to receive public funding for a limited number of foreign students who do not meet these nationality or status criteria, provided they do not account for more than 2% of the total credits for which students are enrolled in the institution or generate a call for more than 2% of the variable component of the public teaching grants (see below).

  3. 3. Students must have sufficient learning credits in their learning credit account (see Box 3.1) to enrol for their selected number of credits. This means, for example, that a student needs at least 30 credits in the learning credit account to enrol full-time for a semester of an associate, bachelor’s or a master’s programme.

In the Flemish funding model, the majority of the lump sum payment for each higher education institution results from the addition of amounts calculated in four distinct calculation components of the model for programmes in university colleges and four main calculation components for universities. University colleges receive a core operating grant that results from the addition of a base component (sokkel) and a variable component for professional programmes (SOWprof and VOWprof) and a base component and a variable component for academic programmes in Schools of Arts (SOWhko and VOWhko). The base components for each institution are calculated using the average number of credits for which students are enrolled for the five-year period t-7/t-6 to t-2/t-2, where “t” is the budget year in question. The base components are designed to guarantee (additional) stability in funding. The amount of funding to be distributed through each base component is established with reference to an initial value fixed in the Flemish Higher Education Code (Flemish Government, 2013[2]) and is adjusted over time for inflation. In 2020, the base components accounted for 10.5% of grant funding for professional programmes and 5.4% of funding for academic programmes in Schools of Arts (see Table 3.1).

The variable components of the teaching grant to university colleges (VOWprof and VOWhko) are calculated with reference to the average values for input and output variables observed in each institution in the same five-year reference period and examined in more detail below. The budget envelope for the variable components is established each year as part of annual Flemish Government budgetary processes, with reference to a provisional budget trajectory established in the Higher Education Code and a statutory adjustment mechanism to take into account changes in enrolment (the “click system”), also discussed below.

The level of the core grant for each university also results from a base and variable component for teaching (SOWun + VOWun), calculated using broadly the same method as for programmes in university colleges, and a based and variable component for research (SOZun + VOZun). Since 2017, programmes in medicine receive a fixed proportion of the variable component of the total university teaching grant (based on their historical share of this budget), with each of the four universities that offer such programmes receiving a share of this fixed proportion of the total budget based on the number of medical degrees they award. In 2020, the base component for teaching accounted for 7.6% of the total teaching grant for universities.

The research element of the operating grant implicitly recognises the strong inter-relationship between teaching and research in research-intensive universities. The level of the specific funding envelope for the research grant to universities is established in relation to the level of teaching grant for universities (SOWun + VOWun) in a ratio of 55% (teaching) to 45% (research). Funding allocated through the teaching grant for academic programmes that transferred from university colleges to universities in 2013 is not taken into account in the calculation of this ratio. The specific budget envelope for research is then distributed to among the five universities using a base component for research (SOZun), calculated based on publication and PhD award numbers, and a variable component for research (VOZun) that takes account of past graduations, bibliometric indicators and indicators related to academic staff. In 2020, the base component accounted for 30% of the value of the research grant to universities (Flemish Government, 2013[2]).

The Flemish funding model is distributive. It uses the calculation components and the parameters discussed in more detail below to distribute the available budget envelope for each of the components among higher education institutions, but does not attach pre-established monetary values to the units (such as credits) that are used in the calculations. The reference value for the budget envelope for each funding component is established in the Higher Education Code, although the Government has latitude to deviate from the amounts established. The result is that the value of each unit of funding distributed by the formula fluctuates over time, depending on the total number of units in the calculation and the total budget envelope available. If the total number of units in the calculation rises – as the number of enrolled credits or credits obtained in the system increases, for example – and the budget envelope remains stable or does not rise in proportion to the increase in units, the value of each unit distributed will fall.

Table 3.1 provides an overview of the specific parameters used to distribute core funding allocated to institutions through the base and variable funding components outlined above. In 2020, almost EUR 1.7 billion was allocated to Flemish higher education institutions through the core model. In addition, as discussed below, various top-up funds were allocated to institutions using funding outside the core funding envelope distributed through the model. For both the teaching and research elements of the grants, the allocation model uses average values for each parameter in the model for the five academic years t-7/t-6 to t-3/t-2, where “t” is the current budgetary year. As such, all funding allocations to institutions are based on the past activity levels and outputs of each higher education institution.

The calculation of the grant allocation for each institution by the Flemish Agency for Higher Education, Adult Education and Qualifications and Student Grants (AHOVOKS) proceeds – in simplified terms – in four conceptual steps:

  1. 1. To calculate the base components for teaching, the number of credits for which eligible students are enrolled in each institution is first adjusted using a degressive weighting system, with higher multipliers for the first tranches of credits (see Table 3.1). This system is designed to provide a proportionally larger base component to smaller institutions. The budget for the three base components for teaching (SOWprof, SOWhko and SOWun), the level of which is fixed with reference to the reference amount in the Higher Education Code, is then divided by the adjusted (weighted) total numbers of credits in each of the programme sectors. Each institution then receives their share of the base component depending on their weighted share of enrolled credits.

  2. 2. The same procedure is applied for the base component of the research grant for universities, using the combined total of PhD awards and publications produced in the system (also weighted for each university using a sliding scale of multipliers) as the denominator to divide the budget for the base component (also EUR 115 million in 2020).

    The divided budget is then allocated among the universities based on their share of weighted PhD awards and publications. The three variable components of the teaching grant are distributed using a system of “funding points”. A funding point is initially generated for each credit for which an eligible students is enrolled, until that student has passed 60 credits in a single degree programme (an input parameter). From the 61st credit a student passes in a single degree programme, a funding point is generated for each credit the student passes. Funding points are also generated for each credit passed by an eligible student under a “credit contract” and for each degree awarded (the latter three variables are output parameters).

    Enrolled credits, credits passed in a degree programme and degrees awarded are multiplied by a factor of 1.5 when they concern students in receipt of a public grant (see Chapter 5), with disabilities, or who combine work and study (werkstudenten). In addition, these same three parameters are adjusted using different multipliers depending on the field of study in which each student is enrolled. As in other OECD jurisdictions, these subject-area weightings (onderwijsbelastingseenheiden – OBE) are designed to reflect cost differences between fields of study and range from 1 to 1.6 for professional programmes and 1 to 4.2 for academic programmes.

  3. 3. Finally, each degree awarded is weighted with a multiplier of 30 for initial professional bachelor’s and master’s degrees and 18 for academic bachelor’s degrees. The budget envelopes for the three variable components of the teaching grant are established in the Higher Education Code, although, as noted, the actual value of the envelope is also influenced by budgetary decisions. These amounts are divided by the sum of the funding points resulting for each component from the multiplications above and then distributed to each institution based on their share of the funding points.

  4. 4. The level of total funding for the base and variable components of the research grant for universities (SOZun + VOZun) is established with reference to the base and variable components of the teaching grant for universities (SOWun + SOZun) to respect a ratio of 45% (research) to 55% (teaching). Funding allocated through the teaching grant for academic programmes that transferred from university colleges to universities in 2013 are not taken into account in the calculation of this ratio. The budget available (VOZun) is then divided by the weighted number of academic bachelor’s and master’s degree awards in each association of university and university colleges; the number of weighted PhD awards; the number of publications and citations calculated for the allocation of the Special Research Funds (BOF – see Chapter 4); and the number of individuals appointed as academic staff for the first time who are female or have moved institutions. The budget for the variable component of the research grant is then distributed to the five universities based on their share of the funding points generated.

To ensure that funding higher education institutions have adequate scale to ensure quality education, the Higher Education Code stipulates that institutions must enrol students in at least 90 000 credits (equivalent to 1 500 full-time students) to be eligible for base funding. This rule does not apply to the specialised Antwerp Maritime Academy. In practice, as discussed in Chapter 2, all higher education institutions in the Flemish Community have enrolment levels well above this minimum threshold.

The current allocation model reflects changes made, from the financial year 2014 onwards, to take account of the transfer of non-arts-related academic programmes from the university colleges to universities. The funding for these programmes is now allocated to universities through the base and variable funding components for academic programmes (SOWun and VOWun). When they were located in university colleges, the programmes were funded using the subject-area weightings for professional programmes (between 1 and 1.6) rather than those used for academic programmes in universities (1 or 2 for the fields in question). To harmonise the system, the weightings for the programmes in question are gradually being adjusted in a process running to 2023. In that year, weightings for architecture or biotechnology will have been moved from 1.46 to 2, while those for languages and management will move from above 1 to 1.

Whereas research capacity associated with most academic programmes in universities is supported through the research element of the core grant to universities, a separate earmarked funding stream outside the model exists to support the development of research in the academic programmes transferred from the university colleges in the form of the “complementary research funds” (aanvullende onderzoeksmiddelen). For reason, as noted above, the proportion of the teaching element of the core grant for universities allocated for “integrated” programmes is not taken into account in the calculation of the size of the research grant using the ration of 55% (teaching) to 45% (research). The complementary research funds are allocated to the “integrated” programmes based on the number of enrolled credits (50%) and the number of degrees awarded (50%) in the programmes concerned. The level of funding for these programmes is allocated based on a planned budgetary “growth path” running to 2023, at which point the intention it to mainstream the funding of research capacity in the programmes into the main funding model.

A more recent change has been the transfer of short-cycle study programmes (graduaatopleidingen) and specific teacher education programmes from adult education centres to university colleges from the academic year 2019/20. These programmes have initially been funded by the transfer of the relevant budget from the adult education centres to the university colleges. In addition, for the years 2020 to 2022, university colleges have been guaranteed additional funds for each additional graduaat student that they enrol through a temporary open-ended funding envelope. The effects of this system will be monitored in 2023 and 2024 and, from 2025, the funding of the graduaat programmes will be fully integrated into the main funding model for the operating grant. From then on, a specific variable component (VOWhbo) will be added to the system, with funding awarded using the same system of funding points as for the other programmes types, albeit initially using a different reference period for the calculation of the parameters (Flemish Parliament, 2018[5]).

The legislation establishing the current funding model for Flemish higher education created a mechanism (the “click system” – kliksysteem) to adjust the level of the budget envelope available for the variable components of the teaching grants to institutions when student numbers in the system increase or decrease. When the number of credits for which students are enrolled increases or decreases by more than 2% in professional, artistic or academic programmes, based on a historical five-year, rolling average, legislation requires the Government to make a corresponding adjustment to the budget envelope for the variable component funding the programme type(s) in question. This system is designed to create a “semi-open” budget envelope, with an automatic, statutory increase (or decrease) in the variable component for any of the programme types when substantial changes in enrolment occur (see Box 3.2). Enrolment in professional and academic programmes has grown steadily since the introduction of the current funding model in 2008/09, triggering regular “clicks” to increase the budget envelope for the variable funding component for these programmes. However, in the years 2015 and 2016, budget pressures meant that the Flemish Government did not implement the budget increases that would have occurred as a result of “clicks” in those years.

In addition to the main lump sum payment stemming from the allocation model and additional funds for associate degree programmes discussed above, higher education institutions in the Flemish Community receive additional earmarked funds, which have been created over time. The most significant of these additional funding programmes are:

  • The “Complementary Research Funds” (Aanvullende onderzoeksmiddelen) already mentioned, which are directed to academic programmes in fields such as architecture that were transferred from university colleges to the universities in 2013 and academic programmes in the Schools of Arts (within university colleges). The purpose of these funds is to support the development of research capacity in these programmes.

  • Funds for practice-oriented scientific research in university colleges (Praktijkgericht wetenschappelijk onderzoek – PWO). This funding, equating to around 3% of the total income of university colleges, is discussed in Chapter 4.

  • Funding for student services (studentenvoorzieningen or STUVO) – the STUVO funds – which support a range of activities, ranging from student housing to guidance and counselling, are discussed in Chapter 5.

  • Funds for capital investment, which are paid to higher education institutions to support investment in buildings and equipment. These funds are additional to the core operating grant to institutions, which, as a block grant, can also be used to fund capital investments, and are theoretically designed to cover investment that support education, research and student services. In 2019, funds for capital investment amounted to around EUR 33 million for the university sector and EUR 23 million for the university college sector – in both cases around 3% of total income from public core funding.

  • Specific funds for initial teacher education (ITE) programmes (Specifieke Lerarenopleidingen – SLO) in both university colleges and universities. These funds, which are intended to support investment in the quality of teaching education, will mostly be mainstreamed into the core funding model from 2022 onwards (Flemish Government, 2013, pp. III.32-33[2]).

  • Additional funds for Brussels-based Flemish institutions. These funds provide extra funding to ensure a diverse provision of a range of Dutch-speaking higher education programmes in Brussels and contribute to higher unit costs of providing certain programmes in Brussels that may not attain the same economies of scale as equivalent programmes in Flanders.

As a basis for reviewing the design and operation of the Flemish funding model for higher education institutions, it is helpful to consider some key criteria that are widely acknowledged to characterise effective funding models. Governments and higher education institutions look for core funding models for higher education institutions to fulfil a range of objectives. Although the degree of emphasis on each criterion may vary between actors in higher education and between countries, the most common objectives for institutional funding models are as follows:

  1. 1. To provide a volume of resources to each higher education provider that is aligned to the real levels of activity, effort and legitimate costs incurred by the institution to ensure adequate levels of quality. Although recognition of effort is important, funding systems that link payments exclusively to activity, such as student enrolment or attainment parameters, can be sensitive to short-term changes in student numbers and risk creating undesirable instability in annual funding levels for institutions with high fixed costs. If not carefully designed and resourced, such systems can also create perverse incentives for institutions to maximise activity (e.g. student enrolment) at the expense of quality or other relevant objectives (Advisory Committee on Funding Higher Education and Research, 2019[6]). Establishing the level of legitimate costs in higher education – and thus the appropriate level of government subsidies to higher education providers – is also challenging (Deloitte Access Economics, 2016[7]). For these reasons, OECD member countries typically combine some form of activity-related resource allocation with other mechanisms to limit funding instability over time and – in some cases – to promote efficiency and quality.

  2. 2. To allocate resources among institutions in a way that is fair, transparent and predictable to support institutional budget planning and ensure the credibility of the system. Maintaining a link between funding and activity, in line with the first objective, and ensuring equitable treatment of similar institutions are key conditions for the fairness of a funding system. It is also important that institutions and their staff understand how budget allocations are calculated and can predict, based on reasonable assumptions, the broad level of funding they will receive in future years to facilitate institutional planning (OECD, 2008[8]).

  3. 3. To support broad access to higher education to help achieve socially desirable skills development and social inclusion goals. Access to higher education depends on a wide range of factors, including the performance of the secondary education system, generic entry requirements and pathways, socio-cultural factors and financial support to students. The design of the funding model for higher education institutions, along with rules relating to admission and student numbers, can also influence institutions’ ability and willingness to enrol (more) students from a wide range of backgrounds.

  4. 4. To promote efficient delivery of good results from education, research and service activities. The results achieved by higher education institutions, including the quality of knowledge and skills acquisition by students, the quality of research output and the efficiency with which students are able to acquire valuable qualifications, are also influenced by a complex range of factors. A level of funding adequate to permit institutions to hire well-qualified, motivated staff and to provide good quality learning and research environments is among the most crucial of these factors. However, human and financial resources can be deployed with different levels of efficiency to achieve the same results. In recent years, governments in many OECD jurisdictions have increasingly sought to use institutional funding models to promote quality and efficiency goals, notably by tying funding to specific outputs and outcomes achieved (de Boer et al., 2015[9]; Ortagus et al., 2020[10]; Zacharewicz et al., 2019[11]).

  5. 5. To permit institutions to invest in strategic development for the future. Like all organisations, higher education institutions need to plan for the future, update their activities and facilities and to shape and react to the changing world around them. Renewal and adaptation require resources in addition to those needed to maintain day-to-day operations. Higher education funding models in some OECD jurisdictions make explicit allocations to support strategic development, either through provisions in core, block grant funding to institutions or through funding targeted through specific programmes.

Combining findings from interviews conducted for this review and comparative analysis of OECD higher education funding models, the following sections examine how the Flemish funding model performs and compares to other OECD systems in relation to the five broad objectives above. These sections analyse, in turn:

  • How the Flemish model of public funding to higher education institutions compares to approaches in other OECD jurisdictions in terms of the structure of core funding appropriations to institutions and the weight of these appropriations in overall institutional revenue.

  • The ways public authorities establish the public budget envelope for higher education institutions, the extent to which budget setting and funding models relate funding levels to activity and ensure fairness, transparency and predictability and how the Flemish funding model compares in this respect.

  • The extent to which the Flemish and other OECD higher education funding models explicitly seek to promote widened access, quality and efficiency, and available evidence on the effectiveness of such approaches – in the Flemish Community and internationally.

  • The mechanisms governments use to support capital investment and future-oriented strategies in higher education institutions and how these compare to the approach adopted in the Flemish Community.

The focus here is on core funding models for operations and instruction in higher education institutions. The specific systems for funding research through institutional funding allocations are examined in more depth in Chapter 4.

In common with a majority of OECD higher education systems, the Flemish Community uses a supplier-oriented mechanism to fund higher education institutions, whereby it provides an operating grant directly to each university and university college. Also in common with its counterparts in other OECD jurisdictions, the Flemish Government regulates the level of tuition fees that institutions can charge, thus influencing the level of resources that institutions can generate from student demand.

In a limited number of OECD jurisdictions, such as England, Australia or Lithuania, governments have explicitly sought to introduce demand-oriented mechanisms for funding higher education institutions, through which public resources are directed to students, who then “spend” these resources on the higher education provider where they enrol. In Lithuania, students meeting merit-based eligibility criteria for state-funded study places are allocated an educational voucher, which they can use to enrol in an accredited higher education institution. England and Australia operate “quasi-voucher” systems, whereby the government provides publicly backed income-contingent loans, which students can use to pay regulated tuition fees in the institution where they enrol. Systems such as Scotland and Ireland operate mixed systems, where public authorities provide direct grants to institutions and pay fees for eligible students that enrol in a given institution.

Figure 3.1 illustrates the average proportion of revenue in universities from core public funding grants, student fees, third parties and other sources in the Flemish Community and five comparable OECD jurisdictions. On average, in 2019, 52% of Flemish universities’ revenue came from institutional grants from government, delivered through the operating grant (around 40% of total funding) and earmarked core funding for research (around 10% of total income from the Special Research Funds (BOF) and the Industrial Research Funds (IOF)). In Ireland and Scotland, where regulated student fees for domestic students and institutional fees for international students are substantially higher than in the other jurisdictions shown, core public funding accounts for only 34% and 39% of average university revenue. In contrast, core public funding accounts for between 57% and 63% of the total revenue of the university sectors in Denmark, the Netherlands and Finland.

Nearly all OECD jurisdictions that provide public funding directly to higher education institutions do so through block grants, allocated as a lump sum that institutions can use largely freely to pay for different expenses. While block grants in many countries are paid to institutions as a single lump sum, they frequently result – as in the Flemish Community – from the addition of different funding components, each with its own calculation basis. The most frequently used funding components are:

  • Basic grants for teaching and institutional operations;

  • Specific grants for research – particularly for research-intensive institutions;

  • Specific grants for capital investment.

In addition, some funding systems make allocations to institutions specifically for strategic investments or policy priorities, such as enhancing quality. This is the case in Finland and the Netherlands, for example.

Table 3.2 presents an overview of the structure of core public funding to both universities and professionally oriented higher education institutions in six comparator jurisdictions, including the Flemish Community as in Figure 3.1. This illustrates that universities of applied science depend to a greater extent on public funds than universities, with core public funding accounting for between 62% (Ireland) and 79% (Finland) of average institutional revenue in non-university sectors. This compares with a proportion of 72% for university colleges in the Flemish Community. Universities in all six jurisdictions receive a research grant as part of their core funding, with the exception of Ireland, where a basic level of funding for research is integrated into the (comparatively small) teaching grant. This is not the case for universities of applied science, which receive no dedicated research grant in Denmark and Ireland and comparatively modest levels of research funding in the Flemish Community and the Netherlands. Only in Finland, does public direct-grant revenue for research make up a substantial portion of average institutional revenue in the university-of-applied-science sector.

Other notable features in the structure of the core public funding models used in these six, relatively similar, systems include a) the practice in both Ireland and Scotland for public authorities to pay fees to institutions for eligible students and b) the existence of dedicated capital grants in three of the six systems. The system of fee payment in Ireland and Scotland reflects the underlying principle in both systems that students are theoretically expected to pay (substantial) fees, but that governments have made a political choice to pay fees on behalf of domestic students.

The tradition of allocating separate grants for capital investment, adopted in the Flemish Community, Ireland and Scotland, is also widespread in other OECD jurisdictions, although the allocation method varies. The OECD Higher Education Policy Survey in 2020 found that, like the Flemish Community and Scotland, the French Community of Belgium, England and Canada allocate recurrent capital grants to higher education institutions (Golden, Troy and Weko, 2021[24]). This is also standard practice for public institutions in many US states (OECD, 2020[25]). In contrast, Ireland’s system of capital grants is based on a competitive application process through which institutions respond to specific calls run by the Higher Education Authority, with similar systems in place in the Czech Republic, Hungary and Poland. Other systems with dedicated capital funding streams, including France, Italy and Norway, award funds based on prioritisation of projects by public authorities. In Denmark, Finland, the Netherlands (see Table 3.2), but also in Austria, Portugal, New Zealand and Sweden, public authorities do not provide dedicated capital funding and higher education institutions are expected to pay for capital investment from core operating grants and/or their own funds (Golden, Troy and Weko, 2021[24]).

Finland, the Netherlands and Scotland allocate a proportion of institutional core funding explicitly for future-oriented investments and strategic development. This issue is discussed in the section below that deals with strategic investment funds.

Higher education funding models in many OECD jurisdictions seek to create a link between the activity levels and costs of higher education providers and the amount of public funding they receive. In doing this, governments typically aim to establish funding models that are fair and transparent and that create a sufficient degree of predictability for higher education institutions. In practice, the extent to which funding models can effectively link funding to costs and guarantee predictability for institutions is affected both by a) the technical design of the allocation mechanism and b) the level of budget envelope that is available to be allocated.

An increasing number of OECD member countries have moved from higher education funding approaches based purely on historical allocations, to formula-based models, using objective parameters to allocate funding to institutions. Of the 27 OECD jurisdictions responding to the 2020 OECD Higher Education Policy Survey, 24 use formulas to allocate at least a proportion of public funding to higher education institutions (Golden, Troy and Weko, 2021[24]).

Along with the Flemish Community, Ireland, the Czech Republic, Hungary, Lithuania, New Zealand, the Slovak Republic and Sweden allocate core public funding to institutions through exclusively formula-based systems. Others systems, including Austria, Denmark, the Netherlands and Norway combine formulas with varying levels of (non-activity-related) “fixed” or historical funding. In Denmark, for example, the level of the basic grant to institutions (grundtilskud), which accounts for 25% of the teaching grant to institutions, is based on historical allocations, as is the “fixed base” (vaste voet) in the Dutch allocation model for the core government grant (rijksbijdragen). Finland is notable for combining a largely formula-based system with funding negotiations for the strategic component of its core funding model (see below).

Nearly all formula systems reported by OECD jurisdictions in the Higher Education Policy Survey use student-related metrics to distribute funds to higher education institutions. Of the 24 systems using formulas, 18 use enrolment metrics (an input parameter) as one of the allocation drivers and a further five, including Finland and Denmark, use exclusively output-based student metrics, such a credits passed and degrees awarded, in their allocation formulas. Among the jurisdictions using enrolment metrics, some, like the Flemish Community, combine both input and output indicators. In all cases, these student-related parameters, which are generally the main driver in formulas, create a link between student-related activity in institutions – and the staff and non-staff costs this generates - and the amount of funding each institution receives.

As discussed below, this does not mean all (or many) systems create a one-to-one link between each enrolled student, each credit enrolled or passed, or each student graduating and a fixed unit of funding. Some funding allocation formulas do use fixed unit prices, usually involving the payment of a fixed sum for each student or graduate in a particular field. Such systems are additive, in that total budget to be allocated is the sum of the unit prices. Many other funding systems – including the one in use in the Flemish Community – use “distributive” formulas, which take the available budget envelope as the starting point and then divide this budget by the number of units that need to be funded. In this case, the value of each unit depends on the number of units and the size of the budget envelope available.

Whether they use additive or distributive formulas, many governments further seek to strengthen the relationship between the notional costs incurred by institutions and the funding allocations they make through incorporating subject-area weightings into their formulas. These provide higher weightings (multipliers) for students enrolled, passing exams or graduating in fields with comparatively high costs, such as laboratory subjects or medicine. Such weightings, while rarely based on detailed assessments of full costs, seek to compensate – at least partially – for differences between fields of study in terms of student-to-staff ratios, technical support staff (such as laboratory assistants) and the costs of infrastructure and equipment (notably laboratories for science subjects).

The Flemish Community and the Netherlands use separate sets of subject-area weightings (referred to in the Flemish context as onderwijsbelastingseenheiden or OBE) for the university and non-university sectors, whereas other systems with binary structures, such as Ireland, Denmark and Finland use a single set of weightings across the whole higher education system. The weightings used in the different jurisdictions are broadly similar, although the range between the lowest and highest subject-area weightings is comparatively small in the Danish system and the difference between the weightings used for universities and university colleges in the Flemish Community is greater than the difference between weightings for universities and universities of applied science in the Netherlands.

Table 3.3 summarises the subject-area weightings used for bachelor’s programmes in the Flemish Community and five comparable OECD jurisdictions. As each jurisdiction uses different subject-area classifications and groupings, the table is designed to provide a broad overview and cannot capture all details of the different systems.

The Flemish funding model allocates higher weights for master’s programmes than for bachelor’s programmes, notably applying a weight of three for master’s in science subjects, compared to two for bachelor’s, and 4 and 4.2 for master’s degrees in medicine and dentistry, respectively, compared to 3.9 for bachelor’s programmes in these fields. The system also applies a number of exceptions to the general rules regarding subject-area weights, for programmes in music and the performing arts, for example, which receive additional funding to reflect their high costs (Flemish Government, 2013[2]). When non-artistic academic programmes were transferred from universities to university colleges in 2013, a progressive system was put in place to align the subject-area weightings for these programme with the weightings used in the university sector by the year 2023. In practice, this has meant increasing the weightings for subjects such as architecture from 1.46 to 2 and reducing the weightings for languages and management from 1.18 and 1.09 to 1. Finally, when associate degree programmes were integrated into the university college sector in 2019, a separate funding mechanism was introduced for these programmes, using subject-area weightings of 1, 1.15 and 1.50 (Flemish Government, 2020[26]). In time, funding for these programmes – and the weightings used for them – will be mainstreamed in the core funding model.

In a recent review of the subject-area weightings and internal funding allocation models used in Flemish universities, de Boer and Jongbloed (2018[27]) found that, while the weightings were unlikely to reflect real cost differences between study fields, the current system was considered fair within the university sector. The same study noted that the Flemish higher education system lacks a standardised activity-based costing (ABC) system that would make it possible to gain a clearer picture of the cost of provision in different fields, as has been achieved in some other OECD systems (KPMG LLP, 2019[28]; Deloitte Access Economics, 2016[7]; Hemelt et al., 2018[29]). As many of the cost studies undertaken in higher education in other countries note, it is questionable whether it is possible to obtain a fully accurate picture of the “true” or “reasonable” costs of providing programmes in a particular field. Not only do costs depend on institutional characteristics and contexts, but a circularity also exists between the revenue higher institutions receive and the amount they are able to spend on particular activities – and thus the cost of these activities in practice (Hemelt et al., 2018[29]). Even if it does not permit judgements to be made about the reasonableness of costs, the transparency created by cost accounting systems nevertheless helps both institutions and governments to understand the observed cost of activities.

Although the legitimate costs of delivering programmes in a given field in a given context are hard to determine entirely objectively, significant misalignment between subject-area weightings and programme costs can be problematic. As noted by institutional representatives in interviews conducted for this review, there has been a growing concern in university colleges about the lack of alignment between the weighting factors used in the Flemish funding model and the real costs of professional programmes. Assumptions about student-to-staff ratios, technical support and facilities that informed the original level of weightings no longer hold, as educational practice has evolved. In particular, there have been concerns about the relative under-funding of programmes in computing and professionally oriented science, technology, engineering and maths (STEM) subjects.

The Flemish Government implemented marginal increases in the subject-area weightings for computing and STEM subjects in university colleges in 2021, acknowledging the need to strengthen advanced education in these fields in the face of growing labour market demand (Flemish Government, 2020[26]). An alternative to this proposal, raised by some interviewees during the review would be to use only one or two subject-area weightings for all types of professional programme, rather than the graduated range of weightings and distinct weightings for associate and bachelor’s degrees used now, with the expectation that institutions will re-allocate funds internally – as many university colleges already do – to address cost differences between fields of study.

A final main way in which governments can seek, through the technical design of their funding models, to compensate higher education institutions for effort and costs associated with delivering education is to provide additional funds for students from particular target groups, who may require additional support. As noted earlier, the Flemish funding model provides an additional weighting factor of 0.5 for students who are in receipt of a government grant (and thus who come from low-income families – see Chapter 5), students with disabilities and students who combine study and work. The intention behind this weighting was both to incentivise institutions to recruit students from these groups – as part of a widening access agenda – and to compensate institutions for the additional costs associated with providing adequate support for these student groups (Flemish Government, 2015[1]).

In practice, as in all distributive funding models, the actual level of additional funding received by institutions for students from target groups depends on the level of the total budget envelope distributed by the model. Hypothetically, if the number of target group students increases equally in all institutions, but the budget envelope for higher education remains constant, no institutions will receive additional funds. If some institutions recruit more additional target group students than others (i.e. increase their share of students from target groups), the model can lead to a (minor) redistribution of the available budget in favour of those institutions with more target group students.

The 2020 OECD Higher Education Policy Survey found that, of the 27 jurisdictions responding, only Ireland, Italy, New Zealand and the Slovak Republic used weightings in their funding models for students from priority groups (Golden, Troy and Weko, 2021[24]). The Irish system, in particular, is similar to the system used in the Flemish Community in that it provides an addition 0.33 weighting for each student from a low-income background; first-time, mature student entrants; students with disabilities and Irish Travellers. As in the Flemish system, the student characteristic multiplier is applied to the weighted subject-related factors, meaning that institutions receive a higher level of additional funding for priority-group students in high-cost subjects. New Zealand uses a multiplier in its formula to direct additional funding for Māori and Pasifika learners, while the system in the Slovak Republic provides funding for students with a Roma background. Other systems, such as those in many Canadian provinces and Norway, direct additional core funding to institutions in regions (such as northern Canada and northern Norway) with large minority populations, but do not use student-related weightings to provide additional funds for students from these groups.

There is a clear logic to providing additional funding to higher education institutions to support the education of students from priority groups for widening access. Previous work by the OECD has highlighted the influence of appropriate guidance and support on progression and completion rates among these groups and the costs associated with providing such support (OECD, 2020[30]). The system in the Flemish Community appears to be supported by institutions and stakeholders. However, as the additional funds generated by the student-related weighting are not earmarked for specific purposes, but incorporated into the operating grant that is provided to institutions as a lump sum, it is not possible to trace with any certainty the ways these additional funds are used and the influence of this funding on institutional behaviour (Flemish Government, 2015[1]). It is nevertheless certain that institutions that enrol large numbers of students in receipt of a government grant (notably in the university college sector) do benefit significantly in financial terms from this mechanism.

While the student-related parameters and cost-related weightings in funding formulas create mechanisms that have the potential to compensate higher education institutions in a proportional and predictable way for effort and costs incurred, the extent to which formula allocation models can achieve this in practice depends on the amount of funding available to allocate.

As noted earlier, some OECD jurisdictions “construct” their budget envelopes for higher education institutions through an essentially bottom-up process. This involves adding together pre-established, fixed units of funding (such as a fixed price per full-time student in a given field, for example) and potential bonus payments, along with allowances for other predictable components of the institutional grant, such as fixed, historical payments. In such systems, governments nearly always calculate the maximum number of students that they will fund in a given year and impose or negotiate enrolment limits for individual institutions or the sector as a whole. As summarised in Table 3.4, this is the system used in Denmark, Australia and Scotland. In systems such as Lithuania – and many other central European countries – the government agrees to fund a specific number of state-funded study places, with students selected on merit (students who do not qualify must pay fees).

Between 2012 and 2018, Australia experimented with a demand-driven system for funding undergraduate education, with fixed payments guaranteed through the Commonwealth Grant Scheme (CGS), but no enrolment caps. The high cost of this programme led to the reintroduction of student limits per institution from 2018 onwards. England has also experimented with a demand-driven system, lifting enrolment caps on institutions from 2015-16 and guaranteeing students admitted to higher education access to public loans to pay the regulated tuition fees. This system remains in place, although the government has considered reintroducing enrolment limits to reduce competition for students between institutions in the system (UK Government, 2020[31]).

Alternatively, governments may fix the funding envelope for higher education institutions in a given budgetary year based on the previous year’s envelope, with adjustments made to account for factors like inflation, projected student enrolment (such as the “reference estimate” process in the Netherlands (Dutch Government, 2021[32])) or political commitments, but without use of fixed or guaranteed units of funding. Funding allocation formulas in such cases are used to distribute the available funds to institutions, with the level of funding per unit in the formula (per student or per credit passed, for example) depending on the total number of units in the formula calculation (the denominator) and the size of the budget envelope (the numerator). As also shown in Table 3.4, the Flemish Community, along with Finland and the Netherlands use this model, albeit with different design features. In the Flemish Community, for example, the level of the budget envelope for the different funding components in the model (the base and variable for the three programme types) is established for up to a ten-year period in the Higher Education Code (Flemish Government, 2013[2]), although the projected budget trajectories are not respected in practice. Ireland uses a hybrid model with fixed unit payments (fees paid by the Government for each student), combined with a distributive formula for the budget envelope remaining after fee payments have been deducted.

The Flemish Community is the only system identified to date in the research for the OECD Resourcing Higher Education Project that combines a distributive funding model with a statutory mechanism to adjust the available budget envelope when enrolment increases or decreases by more than a specified threshold. The “click system” (see Box 3.2) provides a theoretical guarantee that the budget envelope for Flemish higher education institutions will be adjusted in line with enrolment, albeit to a restricted extent. The increase (or decrease) in the budget envelope for the variable components in the funding model in any year is limited to 2%, while an increase (or decrease) in the number of credits in which students are enrolled of at least 2% is required to trigger a “click”. Furthermore, as illustrated by the experience in 2015 and 2016, when budget increases required by the “click system” were not implemented because of budget pressures, government budget setting remains inherently intertwined with the broader fiscal context and political processes, which cannot be controlled fully through regulatory mechanisms.

The base components of the teaching grant in the Flemish funding model are distributed using as the denominator the number of credits for which students are enrolled in the higher education system in the five-year reference period in the relevant programme types. The budget envelope for base components is fixed in the Higher Education Code and has been adjusted over time with inflation (albeit not in line with indicative budgetary planning in the legislation). The allocation model divides the budget envelope for each base component by the number of enrolled credits, effectively creating a payment to institutions per enrolled credit in a particular programme type. The variable components of the teaching grant are distributed based on the number of “funding points” generated by the system in the five-year reference period. Funding points are initially generated for enrolled credits, credits passed and degrees awarded by students eligible for public funding, averaged over the five-year reference period. The multipliers discussed above are then applied to generate a final total number of funding points in the system. As noted, the budget envelope available for each variable component, while nominally fixed in the Higher Education Code, depends on the outcome of a political budgeting process, taking into account the implications of the “click” system. The value of the payment for each funding point depends on the total number of points in each of the three programme sectors in the reference period and the final size of the corresponding budget envelopes.

Figure 3.2 illustrates the evolution of the value of funding points in constant euros (adjusted for inflation) for the three variable components (for professional, artistic and academic programme) of the teaching grant from 2013 to 2021.

The value of a funding point for professional programmes initially decreased, after adjusting for inflation, by 5% between 2013 and 2018, before recovering, resulting in a real-terms decrease of 2% between 2013 and 2021. The value of a funding point for artistic programmes in Schools of Arts declined in real terms by 3.6% over the eight-year period and the value for academic programmes by 10.7%. In the case of the academic programmes, there was sharp decline between 2014 and 2016, followed by a period of relative stability since. As noted, the value of funding points in the variable components of the allocation model depends on a wide variety of factors and it is challenging to isolate the exact causes of the changes in value observed over time. If more students enrol for, or pass, more credits, study in greater numbers in study fields with higher weighting factors or complete more degrees in a given five-year period, the number of units will increase and, if the budget envelope stays constant, the value of each funding point will decrease.

A second way to assess the relationship between the resources allocated through the Flemish funding model and activity in higher education institutions is to examine the core operating grant received per student. Figure 3.3 shows the evolution in the average level of operating grant per unique degree-seeking student in the university and university college sectors between 2015 and 2019, as well as the evolution of the total operating grant for each sector and the number of unique degree-seeking students enrolled. Funding during this period was driven by average levels of student-related input and output parameters in successive five-year periods from 2007-2012 to 2012-2017. Degree-seeking students may study at different intensities (be enrolled for different numbers of credits), so do not all call on the same level of amount of resources from institutions. However, a large proportion of students do study full time and it is reasonable to assume that all students require a basic level of support from administrative and academic staff in institutions, irrespective of their level of study intensity, making funding per unique degree-seeking student a relevant indicator.

For universities (Panel A), the number of unique students with a diploma contract increased between 2015 and 2019 by 5%, the total operating grant from the formula in constant (2015) prices increased by 8%, leading to an increase in operating grant per student in constant prices of 1%. As shown in Figure 3.2, the level of funding per funding point for academic programmes declined by 5% in the same period, a difference possibly explained by an increase in study intensity, graduation rate or enrolment in high-cost programmes in the period. If these variables increase, the number of funding points generated by each student will also increase, meaning the value of a funding point can fall, while funding per student increases.

For universities colleges (Panel B), between 2015 and 2019, the number of unique degree-seeking students increased by 8%, total funding from the formula increased, after adjusting for inflation, by 6%, leading to a real-terms decrease in per-student revenue from the funding model of 2%. The relatively steep increase in enrolment of students with a diploma contract in university colleges between 2018 and 2019 was driven partly by the transfer of associate degree programmes from the adult education centres. These programmes are financed through a separate, earmarked funding stream. For the data presented here, the Government Commissioners counted each student on these programmes for only one-third of a “student with a diploma contract” in 2019, to account for the fact they were only enrolled for four months (Flemish Government, 2020, p. 3[19]). However, the data for the level of operating grant per student in university colleges in 2019 should be interpreted with caution.

International data also suggest there has been a modest decline in real terms spending per student on higher education institutions in the Flemish Community over the last decade. However, international data for the most recent years are not yet available and it is not possible to disaggregate data for universities and university colleges or to isolate institutional revenue from the operating grant (which is reported within an aggregated budget for spending from public sources). Figure 3.4 illustrates the change in public and private spending per full-time equivalent student on higher education institutions in the Flemish Community, the Netherlands and Finland from 2008 through 2017. The Netherlands and Finland were selected as comparators for the Flemish Community as they have similar higher education systems and reliable international data are available for the extended period covered.

Between 2008 and 2017, total public spending (for teaching and research) per FTE student on higher education institutions (universities and non-universities together) decreased, after adjusting for inflation, by 3% in the Netherlands, 4.6% in the Flemish Community and 10% in Finland. In the same period, total private spending per FTE student increased - from a lower base - by 11% in the Netherlands, 24% in the Flemish Community and 67% in Finland. In the Netherlands, this change partly reflects an increase in tuition fees (which are substantially higher than in the Flemish Community), whereas in the Flemish Community the change in private expenditure is likely to have been driven primarily by increased expenditure by businesses and non-profit bodies on research or service activities. The sharp change in private expenditure in Finland, albeit at a modest level, may partly reflect a change in institutional revenue classification with the change of the ISCED system in 2012.

The same international data on public spending per student on higher education institutions can be disaggregated into spending destined for R&D and spending for core and ancillary services (which includes teaching and operations), albeit only for the years since the 2012 ISCED reform. These data show that, while private spending on higher education institutions in the Flemish Community increased by 15% and public spending on research by 12% in real terms between 2012 and 2017, public spending on core and ancillary services decreased by 3%, after adjusting for inflation. In the Netherlands, Sweden, Germany and Finland public spending per FTE student on core and ancillary services decreased by more than 3% in real terms and public spending per student research also fell, with the exception of Germany.

Drawing together the various elements discussed above, it is possible to draw some interim conclusions about the extent to which the Flemish funding model for higher education institutions is fair, transparent and predictable.

Fairness implies both that institutions are treated equally and that they receive their “fair share” of funding available in light of their real level of efforts, activities and costs. On balance, the Flemish funding system can be considered fair, as allocation of funds is based on objective criteria, which apply to all publicly funded institutions, and the model explicitly seeks to account for differences in the cost of providing quality education between programmes and institutions. The design of base funding components provides proportionally higher levels of funding to smaller institutions, compensating them partially for their more limited ability to generate economies of scale. The subject-area weightings, while partially arbitrary, seek to compensate institutions for differences in the cost of provision between fields and the weighting for students from designated target groups provides addition resources to devote to students who are typically most in need of support. The use of historical student-related data (t-7/t-6 to t-3/t-2) for the calculation of the allocations reduces year-on-year fluctuations in institutional funding, but weakens the alignment between activity levels and the funding received in any given year, which broadly disadvantages institutions with expanding enrolment and protects institutions with declining enrolment.

Whether or not the system is transparent depends on the perspective from which is it viewed. With modest application, it is relatively straightforward to understand the calculations that are made to allocate the available budget envelopes to institutions. However, the relatively large number of parameters and weightings mean that the Flemish funding model is undeniably complex. Models in comparable OECD jurisdictions use fewer parameters and weightings, which generally makes them easier for an average citizen, student or academic staff member to understand, but does not mean they are more effective. The complexity of the Flemish model results in large measure from the efforts to ensure institutions receive appropriate compensation for their efforts, as identified in the point above. However, the coexistence of multiple parameters in a distributive allocation model means the relationship between the value of a given activity or output parameter and the financial consequences for the institution is not always clear. Moreover, the real financial effect of many of the parameters used is further diluted by the use of historical averages as a calculation basis. In this respect, additive funding models, which provide fixed unit payments for different parameter values in the previous year, create a more direct relationship between activity or performance and the funding institutions receive. However, as noted above, such additive models are usually applied in combination with other mechanisms (such as student caps) to limit the total call on public budgets, while use of a single year as a calculation basis can result in larger year-on-year fluctuations in funding for individual institutions.

The funding model does not result in major annual changes in institutional funding, but neither does it guarantee a consistent value for the units of funding institutions will receive over time. The use of the five-year average of student-related variables as a calculation basis for the funding allocation and legal safeguards limit the scale of changes from one year to the next year in the overall level of funding institutions receive. However, the moving variables in the system that influence the number of funding points in the variable components of the model (enrolments, subject choice, study intensity, progression and completion) are influenced by the activities of all higher education institutions offering programmes of the same type (professional, artistic, academic). As such, institutions must take into account the activity of all institutions offering the same types of programme to be able to calculate and predict their share of the total budget envelope. However, this task has been complicated further as the level of funding in the budget envelopes to be distributed through the formula has frequently not been adjusted in line with the indexation method and budget trajectory specified in the Higher Education Code or, in some years, the “clicks” that were triggered by increased enrolment. This unpredictability of the number of units by which the budget envelope must be divided and of the size of the budget envelope itself, make it challenging for institutions to plan their budgets precisely.

Achieving the appropriate balance between stability and proportionality of funding to current cost levels is inherently challenging. However, as discussed in the conclusions to this chapter, there is almost certainly scope to amend the approach to setting budget envelopes and simplify elements of the current model to create a more transparent and predictable system.

A frequent criticism of the system heard during the review interviews was that the current system creates a “zero-sum game” in the higher education system – and particularly among university colleges - as institutions compete for students. In reality, however, this issue does not result from the design of the funding model or even the level of funding in the system. Rather, it is the product of a largely stable student cohort and the existence of multiple institutions in the university college sector with similar profiles. Even if the system guaranteed a fixed amount of funding per student in a given field and funding payments kept pace with inflation, institutions offering the same programmes would still need to compete for students in a competitive market place. As discussed later, short of imposing recruitment limits on institutions – as done in some OECD systems – the solution to this issue in the Flemish context is most likely to lie in some further sharpening of the profiles of institutions that compete directly with each other for students, to allow them to offer more distinctive and differentiated programme mixes.

Whereas the fixed component of the teaching grant for Flemish higher education institutions is allocated based on the number of credits for which students are enrolled (a measure of input), the variable component of the model – which allocates over 90% of the total teaching grant – uses both input and output parameters. Until students pass their first 60 credits in a single degree programme, institutions receive funding for them from the variable funding components based on the number of credits in which the students in question are enrolled. This system is designed to compensate institutions for the costs they incur in educating and orientating students in the first stage of their higher education career. By rewarding enrolment in – rather than passing of – credits, the model guarantees institutions funding even if students initially fail courses or need to change programme. This is an important consideration in higher education systems with a comparatively low bar to entry, as in the Flemish Community, as there is typically a greater need to orient and support students than in selective systems. As discussed in the previous section, as with the other parameters in the variable component of the model, enrolled credits are weighted to account for cost differences between fields of study and to provide additional funding for students from non-traditional student populations.

From the point that students pass their 61st credit in a single degree programme, the funding model compensates the host institution for each credit the student passes and, ultimately, for the degree (diploma) that they obtain at the end of the programme. Students who have already successfully passed 60 credits still count towards the calculation of enrolled credits for the base component (sokkel) in the funding model. The output parameters were included in the model in the 2008 reform to reward institutions for supporting students to complete courses and programmes successfully (Flemish Government, 2015[1]).

As discussed, a majority of OECD jurisdictions use formula and indicator-based models to allocate at least a proportion of the public funding they provide to higher education institutions. Among these systems, an increasing number have included output and – in fewer cases – outcome variables in their formulas. Among the 27 jurisdictions responding to the 2020 Higher Education Policy Survey, 24, including the Flemish Community, indicated that they use a formula-based model to allocate funding to higher education institutions. Figure 3.6 presents the input, output and outcome variables that these jurisdictions indicated that they use in their formulas for core institutional funding (which may include both teaching and research grants). While enrolment – an input variable – is the parameter used in the largest number of systems, research outputs (number of publications and number of PhDs awarded) are used in around half the systems covered, while degrees awarded and credits completed – measures of educational output – are used in 11 and 8 systems, respectively.

Outcome indicators related to graduate employment (such as graduate employment rates) are used in 10 systems, but few systems go as far as to link funding to graduate employment in sectors related to their field of study or graduate earnings.

A majority of OECD jurisdictions that allocate funding based on indicators do so through formula-based allocation models, as is the case in the Flemish Community. However, in a smaller number of systems, indicators are used in funding allocation in two additional ways, as summarised in Table 3.5. The first of these is to tie the award of funds from an additional “pot” of money to achievement of system-wide performance goals. In Denmark, for example, 7.5% of the budget envelope for the teaching grant to higher education institutions is allocated outside the main output-driven model, based on institutions’ performance in relation to average study duration and the employment rate of graduates.

The second method is to tie the payment of a specific proportion of the institutional funding envelope to institutions’ performance in relation to quantitative targets established in institutional performance agreements. The first generation of performance agreements in the Netherlands (between 2012 and 2017) used a set of seven nationally determined indicators, for which institutions agreed their own targets. Payment of 5% of the total teaching grant was made dependent on achievement of these quantitative targets. The system of institutional compacts in Ireland establishes a long list of indicators at national level, from which institutions can choose in formulating their own institutional targets. Under the first generation of these institutional compacts (2014-17), payment of up to 10% of the total teaching grant was made conditional on progress towards targets established by institutions, as part of a regular process of monitoring and dialogue between institutions and public authorities.

The new system of quality agreements in the Netherlands, running from 2019 to 2024, which followed on from the previous experiment, makes payment of a bonus funding allocation in 2024 dependent on institutions’ making satisfactory progress in relation to the goals of their quality agreements. However, rather than focusing primarily on centrally determined quantitative metrics, it relies on qualitative assessment by the national accreditation body (the Accreditation Organisation of the Netherlands and Flanders – NVAO). The financial penalties provided for with the Irish system of institutional performance compacts were never applied in practice. Under the Irish system performance framework for 2018 to 2020, a modest level of additional funds have been made available annually by the Government, with competitive financial awards made each year to institutions that can demonstrate particularly good examples of achieving objectives established in their compacts. These awards are made on the basis of qualitative case studies, submitted by institutions and assessed by an independent panel (HEA, 2019[33]).

Table 3.6 provides an overview of the proportion of the teaching grant to higher education institutions in six OECD jurisdictions, including the Flemish Community, that is allocated using input, output and outcome variables, as well as historical allocations that are not determined by indicators. This illustrates that the Flemish Community allocates a higher proportion of funding for teaching based on output variables than comparable jurisdictions such as Estonia, Norway or Ireland (the latter using a formula driven by enrolment numbers), a similar proportion to Denmark and a smaller proportion than Finland. Denmark and Finland both link a small proportion of total funding for teaching to outcome indicators: graduate employment outcomes in Denmark and the results of student feedback surveys in both Denmark and Finland.

Although an increasing number of OECD member countries have introduced output and outcome-related funding models, robust research into the effects of such systems has been limited. State governments in the United States were among the first in the OECD to embrace output-based funding, initially in the 1980s and 1990s and subsequently in another wave of reforms in the 2000s. As a result of this early experimentation, as well as the advanced evaluation capacity that exists in the US scientific community, most available studies into the effects of performance funding are from the United States. A significant number of these deployed robust quasi-experimental research designs. A recent analysis of the results of these studies (see Box 3.3) found only limited evidence of positive effects from performance-based funding systems on target variables, such as student progression and completion rates. The analysis also found widespread examples of unintended and undesirable consequences (Ortagus et al., 2020[10]).

Fewer studies have investigated the impact of output and outcome funding in European higher education systems, although the evidence that does exist suggests a similarly limited impact. A study in Denmark found the completion-oriented “taximeter” system to have had a mixed influence on completion rates in Danish higher education institutions. At the Copenhagen Business School, for example, the implementation of the taximeter was followed by an increase in completion rates at the bachelor’s level, but a reduction in rates at the master’s level (Claeys-Kulik and Estermann, 2015[34]). Likewise, an evaluation of different performance-based funding formulas used in German federal states between 2000 and 2008 found that their introduction was rarely followed by significant changes in the outcomes they sought to influence, casting doubt on their efficacy, particularly given the cost of their implementation (Dohmen, 2016[35]).

In contrast to the limited effects found from the output-based formula systems, the study of German higher education systems by Dohmen (2016[35]) found that “target agreements” (Zielvereinbarungen), in which institutions specify goals and actions and agree these with government in exchange for funding, were associated with more positive effects. Notwithstanding the challenges of proving causality, these reported effects included observable changes in measurable indicators, such as increases in third-party funding and improved graduation rates in universities of applied science. Perhaps more significantly, the introduction of performance agreements in German federal states was found to have led to an increased focus on results and more strategic, evidence-based decision-making in higher education institutions. This is consistent with analysis of the introduction of performance agreements in Finland, which is reported to have increased understanding and management of costs and the focus on performance within universities. Another study, in North-Rhine Westphalia (Germany), reported in de Boer et al. (2015[9]), also found that performance agreements provided a basis for better internal decision-making in higher education institutions. A similar pattern was found in Ireland in relation to the system of institutional compacts, which appears to have had limited direct effect on the behaviour of institutional staff and observed outputs, but to have improved institutional strategy and dialogue between the institutions and public authorities (O Shea and O Hara, 2020[36]).

The systematic evaluation of the first generation of Dutch performance agreements also concluded that the agreements had generated positive effects on the organisation and strategic focus of higher education institutions (Reviewcommissie Hoger Onderwijs en Onderzoek, 2017[37]). In particular, the review commission responsible for the evaluation argued that the process of developing, negotiating and monitoring the agreements had helped higher education institutions to refine their institutional strategies, tailor their educational offerings and, in universities, sharpen their research profiles. The evaluation also noted that pass rates and on-time completion rates in universities increased during the implementation period for the performance agreements, but that on-time completion rates in bachelor’s programmes in universities of applied science actually decreased (from 70% to 67% overall), particularly in the large institutions in the Randstad. The review team acknowledged that the inherently challenging (or impossible) task of establishing causal relationships (either positive or negative) between the performance agreement system and outputs (such as pass rates) was made even harder by an accumulation of other policy changes that were implemented in parallel.

The Dutch review commission examining the first generation of Dutch performance agreements concluded that a new generation of agreements should avoid the strong focus on centrally determined quantitative indicators and adopt a more qualitative approach, albeit with measurable indicators of progress at institutional level:

The committee recognises the limitations of working with indicators: not everything that is valuable can be measured. It is therefore important that in the assessment of and accountability for the agreements there is room for the context and the underlying story of the institution. Performance agreements offer the possibility of a strategic dialogue with the institution. The risk of strategic behaviour and perverse effects is greater if performance indicators are part of a mechanically applied formula in the funding model. The committee recognises the importance of qualitative goals, but is of the opinion that there must also be demonstrable efforts and results.[OECD translation] (Reviewcommissie Hoger Onderwijs en Onderzoek, 2017, p. 73[37])

The balance of evidence internationally therefore suggests that performance agreements (or “quality agreements” in their latest iteration in the Netherlands) can have positive effects on system governance and institutional strategy, but that their impact of core output variables is likely to be limited or, at best, mixed. This raises the question of how to formulate the objectives of such agreements in a realistic way, how to incorporate specific, measurable, achievable, relevant and time-bound (SMART) measures into agreements and whether and how to link agreements and institutional funding. It seems likely that institutional agreements might function most effectively as accountability, transparency and strategic planning tools and that these should be their primary objective. To function as accountability tools, to take up the phrase of the Dutch review commission, there must be “demonstrable efforts and results”, but the use of one-size-fits-all indicators is ineffective as it masks complex realities. Using tailored institutional agreements with a limited number of SMART targets that can be assessed through qualitative as well as quantitative methods could be a promising approach. This is broadly the model now adopted in the Netherlands and Finland.

In terms of linking funding to performance agreements, different models exist as illustrated in Table 3.5. In all cases, with the exception of Ireland, the proportion of funding made dependent on achieving goals in the institutional agreement is small. In Ireland, the nominal rate of 10% of total funding linked to institutional compacts is largely notional, as no institutions have ever incurred a financial penalty in practice (O Shea and O Hara, 2020[36]). The limited available evidence suggests that attaching a small amount of money to institutional agreements can be sufficient to incentivise institutions to take the agreements and the process seriously, while avoiding perverse effects that might arise from a process with higher stakes in financial terms (de Boer et al., 2015[9]; Reviewcommissie Hoger Onderwijs en Onderzoek, 2017[37]). As with most public policies, introducing systems of performance agreements is best implemented with the addition of at least some new funds to the overall funding envelope for higher education institutions, rather than purely through the reallocation of existing funds. For this reason, the Irish Higher Education Authority (HEA) has moved to a system that provides additional “bonus” payments to institutions that can demonstrate particularly strong examples of achievement of goals in their institutional performance compacts (HEA, 2019[33]).

The output-linked funding model introduced in 2008 in the Flemish Community has not been subject to the kind of robust impact analysis used in some of the US studies analysed by Ortagus.et.al. (2020[10]) and highlighted above. The internal evaluation undertaken by the Flemish Government in 2015 came too early to observe long-term trends, and other reforms, including the recent transfer of academic programmes out of university colleges, complicated the analysis (Flemish Government, 2015[1]). Nevertheless, the 2015 evaluation already noted that there was no evidence of an improvement in progression and completion rates in period between 2008 and 2015 and that average “time-to-degree” had even increased.

As shown in Figure 3.7, the rate at which first-time students in bachelor’s programmes drop out of their studies after one, two and three years have remained stubbornly constant – and even slightly increased in the period since 2008. Other evidence suggests that the proportion of degree-seeking students enrolling for less than the full-time equivalent of 60 credits has risen in recent years – for example from 16% to 18% between 2016/17 and 2017/18 (Flemish Parliament, 2019[39]).

Apart from the inherent challenges of supporting students from an increasingly diverse set of backgrounds to choose study programmes aligned with their capabilities and work effectively to pass their courses in a timely manner, several aspects of the Flemish funding and study credit model are likely to influence progression and completion rates. Firstly, the incentives for institutions to adapt their behaviour in response to the inclusion of output parameters in the funding model are weak. In particular, there is no limit on the time that a student can take to acquire their first 60 credits, up to which point institutions are funded on the basis on enrolled credits (inputs).

Perhaps more significantly, the structural mechanisms designed to incentivise students to make wise choices about what to study and to progress swiftly through their studies also appear to fail to create strong incentives. The learning credit (leerkrediet) system (see Box 3.1) was conceived as a way to influence student behaviour by limiting the time they could spend in higher education without progressing successfully. However, the high starting credit of 140 credits and the 60-credit bonus students receive on passing their first 60 credits mean that students often have “spare” credit in their account and may have limited additional incentives to be careful about study choices, study more intensively or progress more swiftly through their degree programme.

At institutional level, a legally regulated system of “study progress monitoring” (Studievoortgangsbewaking) is in place, which allows institutions to impose mandatory conditions on students who do pass sufficient credits and, ultimately, to refuse to re-enrol them if they continue to show poor progress. A review of this system, which also considered the learning credit mechanism, concluded that the study progress monitoring process intervened too late, when students had failed 50% of their credits (Werkgroep "Studievoortgangbewaking", 2014[4]). It recommended a lower threshold for intervention at institutional level and a “tightening” of the study credit system at system level. The study recommended removing the double counting of the first 60 credits passed by students and placing limits on the time students can take to obtain a qualification (time to graduation). Of these recommendations, only the change to the intervention threshold for study progress monitoring appears to have been implemented (intervention now occurs if students fail 40% of the credits for which they are enrolled).

There is evidence from Flemish higher education and internationally that students’ initial momentum in their studies has a determining impact on their chances of success. Students in Flemish higher education who enrol for higher numbers of credits in the early stages of their higher education career are found to have higher chances of passing credits and obtaining qualifications than students who take lower numbers of credits (Werkgroep "Studievoortgangbewaking", 2014[4]). This finding is supported by a body of work in the United States focusing on study “momentum”, which also finds that students with higher study intensity at the start of their higher education career are significantly more likely to complete a degree (Attewell, Heil and Reisel, 2012[40]; Clovis and Chang, 2019[41]). As it is understood that the proportion of first-time students in Flemish higher education enrolling for fewer than 60 credits has increased in recent years, this is a factor that warrants careful consideration in policy design.

Alongside ensuring that higher education institutions have adequate resources to operate on a day-to-day basis and that institutions fulfil their missions efficiently and effectively, governments have an interest in making sure that institutions can invest in their longer-term strategic development. In this regard, two, inter-related, types of investment are likely to be of greatest importance for the long-term sustainability of higher education institutions.

The first is investment in buildings and equipment (capital investment), which intrinsically requires a longer planning and budgetary horizon than expenditure on operations, and typically involves large volumes of resources. The second is investment in institutional changes to respond to an evolving environment and to improve the effectiveness of activities. These two types of investment are inter-related because decisions about long-term investments in capital need to be informed by an understanding of future organisational structures and approaches to learning, research and service. The design of buildings, for example, needs to take into account factors such as the predicted future size of the student and staff bodies, how staff and students will work together and, particularly in light of increasing digitalisation, the balance between on-campus and online learning.

Chapter 7 of the report considers some of the main areas where innovation and structural change are likely to be required in higher education systems – including the Flemish higher education system – in coming years. The remainder of this section considers how higher education institutions obtain and allocate revenue for strategic investment in capital and innovation in the Flemish Community and in other OECD jurisdictions.

Higher education institutions require good quality, fit-for-purpose buildings and equipment to fulfil their missions to a high standard. Moreover, the quality of buildings and equipment plays an important role in institutions’ ability to attract staff, students and external and contract funding. In analysis of capital funding in the UK, Frontier Economics (2015[42]) based their analysis on the assumption that an institutional operating surplus of 7% is required simply to maintain buildings and equipment in good order. Additional resources are required to invest in expansion, major reconfigurations or replacement of existing building stock.

As noted earlier in the chapter, some OECD jurisdictions, including the Flemish Community, Ireland, England and many US states, provide dedicated capital grants to higher education institutions. This earmarked funding is additional to the core block grant funding, which institutions in many systems, including the Flemish Community, are free to spend as they wish, including on capital projects. Other OECD higher education funding systems, such as Denmark and Finland, provide no dedicated capital funding and expect higher education institutions to plan and finance their investment through the block grant funding they receive from the state and their own funds. There are advantages and disadvantages to both these approaches. A dedicated capital grant makes explicit the need to invest in buildings and infrastructure and can create an opportunity for serious discussions on how much funding is required for a long-term capital programme. The experience of several OECD jurisdictions with dedicated capital grants, including in all the jurisdictions cited above, is that they are typically insufficient to cover the capital investment needs of institutions. Moreover, given the less immediate impact of funding reductions, tend to be a target for cuts in times of fiscal constraint. The level of capital grants also tends to follow the availability of funding at government level, rather than the timing of needs within institutions, creating the risk that funds are unavailable when most needed or poorly invested when available.

Relying on core funding to fund capital projects means that institutions have the autonomy and flexibility to plan their own capital investments in line with their institutional strategies and incentivises institutions to get the best possible value for money out of the investments they make. However, there is always a risk that capital investment is “forgotten” when determining the size of budget envelopes for higher education institutions at system level and that capital investment is seen as a luxury, rather than a necessity.

In the Flemish Community, universities and university colleges fund the lion’s share of their capital investments with their own reserves and the level of borrowing to fund investment is reported to remain limited. In 2019, capital grants were estimated to cover only 6% of the investments made by Flemish universities (Flemish Government, 2020[14]). The interviews conducted for this review revealed that the low level of public capital subsidies has not stopped Flemish higher education institutions form investing in new buildings and renovations. However, higher education institutions report that there is a backlog of maintenance and investment in the system and that a large proportion of historical buildings and – often poor quality – buildings dating from the large-scale expansion of higher education in the 1960s and 1970s will soon require major renovations or total replacement. This is all the more important given the Government’s emissions targets and new regulations on the energy efficiency of building stock. The challenges of adapting the system to digitalisation and lifelong learning add to these underlying issues.

Higher education institutions typically fund some investment in strategic development from general operating revenue. However, major new projects, institutional restructuring and significant investments in innovation, whether in digital infrastructure, staff development or new laboratory equipment, require investment capital. Governments explicitly or implicitly provide such resources for investment in three main ways:

  1. 1. In many OECD higher education systems, there is a general expectation that institutions will fund strategic development from their own funds, a large proportion of which may come from public core funding that is nominally intended for day-to-day instruction or research activities. Institutions are typically able to re-allocate funds internally and to create internal funds for strategic development at institutional or departmental level. This is the broad expectation in systems such as Denmark, the Flemish Community of Belgium and, historically, the Netherlands.

  2. 2. A second, frequently used, approach is for governments to provide targeted funding to institutions for strategic development or innovations in specific areas linked to government priorities as earmarked grants or through competitive calls for proposal. The vast majority of the 28 OECD jurisdictions responding to the Higher Education Policy Survey on resourcing indicated that they used targeted funding for strategic priorities. This is the approach used in Ireland, for example, where a proportion of the total budget envelope for higher education institutions is reserved (“top-sliced”) for national targeted funding programmes. Although targeted funding allows governments to ensure resources are allocated to policy priorities, the allocation of funds typically occurs through calls for proposals, which can be time-consuming and administratively burdensome. If the sums of money involved are relatively small, such calls can quickly become inefficient for both public authorities and institutions.

  3. 3. A final approach is for governments to explicitly allocate a proportion of core funding to institutions for strategic development. In practice, in the primary examples of this approach identified in the OECD Resourcing Higher Education Project, the funds for strategic development are added to the block grant to institutions to allocate freely, but the strategic component of the funding is explicit, rather than implicit as in the first approach. Finland and Austria both use institutional funding models with three pillars for a) education; b) research and c) strategic development. A majority of funds for strategic development in Finland and all such funds in Austria are allocated to institutions as part of the block grant, but institutions have to account for their use of the funds through performance agreements. The Netherlands uses a similar approach, although the strategic funds take the form of a dedicated quality fund (kwaliteitsbekostiging) that is earmarked for a broad set of quality-related investments (Dutch Government, n.d.[43]). Here again, the funds are intended to support activities and objectives agreed in quality agreements (see above) between the government and each beneficiary institution.

The need for investment in strategic development emerged as an important theme in the interviews with higher education institutions conducted for this review. The high proportion of institutional revenue absorbed by staff and fixed costs in the Flemish Community leave institutions with little margin to invest in future-oriented developments.

The analysis in the preceding sections has highlighted the careful design of the core funding model for higher education institutions in the Flemish Community, with its focus on ensuring a fair allocation of funds and on rewarding institutions for supporting students to progress within, and complete, their studies. Nevertheless, the analysis also pinpoints a number of challenges related to the design and operation of the current system:

  1. 1. In operation, the funding system has only succeeded in creating a comparatively weak alignment between real activity and cost levels in higher education institutions – generated by changes in enrolment, for example – and the amount of funding these institutions receive. In recent years, the value of each “funding point” in the variable component of the model has declined in real terms in universities and stagnated in university colleges. This affects institutions’ ability to plan effectively and invest in quality education.

  2. 2. Including output-related parameters in the funding model, alongside input indicators, has not resulted in the desired changes to observed outputs and has proven a largely ineffective way for government to influence performance in the system.

  3. 3. The current system of core funding and associated supplementary grants allows institutions freedom to allocate funds internally, but, in practice, provides limited resources for institutions to invest in strategic development, including future-oriented capital investment.

The remainder of this section examines possible policy options to address these three issues.

The core teaching grant for publicly funded higher education institutions in the Flemish Community is composed of two main types of unit: an amount in euros for each credit for which students are enrolled from the base component of the model and an amount in euros for each funding point generated by the variable component of the model. The value of these units depends on:

  1. 1. The average total number of study credits for which all eligible students in Flemish higher education institutions are enrolled for the five-year period t-7/t-6 to t-3/t-2 and the distribution of this enrolment between smaller and larger institutions (see Table 3.1). These factors drive the value of the base component of institutional funding.

  2. 2. The average number of funding points generated by student-related input and output variables and multipliers in the same historical five-year period. This drives the value of the variable component of the grant.

  3. 3. The value of the budget envelopes available for distribution through the base and variable components in a given budget year.

The first and second variables will always change for year to year, as students’ enrolment decisions and behaviours change. As discussed earlier, the size of the budget envelopes available depends on the nominal annual budget trajectory established in advance the Higher Education Code, the level of annual indexation to account for inflation, the influence of the “click” system and – crucially – political choices on whether and how to implement the budget allocations and adjustments provided for in the relevant legislation. The combination of these moving variables makes it challenging for institutions to calculate their annual budget in advance. The student-related variables result from the total level of activity in all programmes of the same category (professional, academic, etc.) in the Flemish Community, not just in a single institution, while government decisions about the size of the total budget envelope have a fundamental impact on the level of funding institutions receive, with some changes occurring at short notice.

The higher education sector in the Flemish Community stresses, in particular, that the Government has not increased the budget envelopes in line with inflation (indexation) and the planned budget trajectories (“growth paths”) specified in the Higher Education Code in recent years. Moreover, governments have repeatedly failed to implement upwards “clicks” to the budget envelope triggered by enrolment increases. This occurred most recently in 2015 and 2016. The current government (2021) also plans to implement smaller increases to the budget envelope for higher education institutions than those set out in the budget trajectory published in the Higher Education Code and not to implement “clicks” triggered by enrolment growth for academic programmes in universities for the years 2021 and 2023 (Flemish Government, 2020[44]).

Increasing the budget envelopes would be the least technically challenging way to address the real-terms decline in the value of funding points for academic and artistic programmes discussed earlier. Respecting the budget trajectories and principles for indexation and “clicks” established in advance in legislation would also significantly increase the transparency and predictability of the funding system for institutions.

However, increasing the budget envelope in line with planned budget trajectories alone would not change the fact that the funding model, in combination with the open access admissions system, allows unit funding rates (resources per funding point) to fluctuate over time. Fully eliminating such fluctuation would almost certainly require the creation of fixed unit funding rates, such as fixed payments for specific groups of study field for each credit that students enrol for and pass, as used in some other OECD jurisdictions. However, if the current open access admission system in Flemish higher education – with no enrolment caps – were maintained, a system of fixed payments would require a move to an “open” budget envelope, as the cumulative value of the unit payments required to compensate institutions for student-related activity could easily exceed any planned “closed” budget envelope.

The only notable recent experience of an OECD jurisdiction using an open budget envelope was Australia’s experimentation with “demand-driven” enrolment in the years up to 2018. This experiment was halted because of budget constraints, with the introduction of enrolment caps at institutional level. As the Flemish system allocates funding based on historical enrolment, progression and graduation trends, rather than simply the previous years’ enrolment as in the Australian system, public authorities would have a greater ability to predict budgetary requirements and calculate unit payment levels in advance. Denmark applies a funding system with regularly adjusted unit payments that are established for several budget years at a time. However it uses a funding formula with fewer parameters than the Flemish system and overall enrolment limits for higher education institutions. For universities, for example, in each annual budget round, the Danish Government calculates fixed unit prices per 60 passed credits (student full-time equivalent or studenterårsværk) in three subject price categories for the next three years, based on past completion rates and enrolment projections.

“Additive” funding models, with fixed unit payments established for a number of years in advance, can be simpler to communicate than the current model and create a more transparent relationship between institutional actions (such as recruiting a student from a target group or successfully supporting a student to acquire a degree) and the amount funding the institution receives. A fixed unit payment for students from specific target groups effectively guarantees institutions additional funds for each student from these groups, unlike the multiplicative weightings as used at present in the Flemish Community and in Ireland.

However, despite these advantages, it is questionable whether an additive model could be made to be financially sustainable in a higher education system with an open access admissions model system, such as that in the Flemish Community. Particularly in the current economic climate, the financial implications of such a model are unlikely to be palatable for the Flemish Government. Nevertheless, given the concerns about ensuring transparency and simplicity in funding in parts of the Flemish higher education sector, it would be instructive to analyse the effects of such an additive model, both on public funding requirements and on the distribution of funds between institutions.

Any move to a funding system with fixed unit payments would not fundamentally change the reality of intense competition between institutions for students – an issue frequently raised by interviewees consulted during this review, particularly in the university college sector. When the funding an institution receives depends largely on student-related parameters, there will always be a strong incentive for institutions to recruit as many students as feasible, including through competitive recruitment practices. In the Flemish context, this competition is almost certainly exacerbated by the existence of institutions offering similar programmes in close geographical proximity to each other. While some higher education systems, such as England, have historically tackled the risk of excessive inter-institutional competition by imposing enrolment limits for each institution, such an approach is incompatible with the open access model in the Flemish Community. A possibly more promising approach might be to support greater institutional specialisation. This topic is addressed in Chapter 7.

OECD member countries use multiple policy tools in their efforts to ensure higher education institutions perform their activities to a high quality and make efficient use of resources. Among these tools are regulatory measures relating to the qualification of staff, external accreditation and quality assurance systems, financial audits, systems of transparency and reporting and public funding systems. As discussed in this chapter, research in different OECD jurisdictions suggests that performance-linked formula funding models tend to have limited effects on institutional performance and may generate perverse effects. In the Flemish Community, there is no evidence of the current funding model for the teaching grant, which includes credits passed and degrees awarded as output parameters, leading to unintended consequences, such as a lowering of standards or increased selection in institutions. Equally, however, there is no robust evidence of the model having positive effects on the variables it sought to influence: student progression, completion rates and time to degree.

Given that the inclusion of output indicators in the funding model appears to do no harm and highlights the importance of progression and completion as policy objectives, there is no pressing reason to eliminate these indicators from the funding model. Moreover, a radical change to the parameter mix in the funding model risks creating instability in the system for no clear benefit. Nevertheless, it would be unwise to assume that the inclusion of output parameters in the funding formula will have a significant impact on progression and completion rates in the years to come.

The commitment in the Flemish Community to allowing open access to higher education means that the system will always have to deal with higher rates of initial re-orientation of students, slow progression and student drop-out than systems that select students rigorously at entry. Nevertheless, there may be scope to increase the incentives for students to progress effectively in their studies by adjusting the systems of “learning credit” and study progress monitoring, while maintaining and developing existing and valuable orientation and support services for students as they enter and advance within higher education.

Alongside such actions, there is also scope in the Flemish higher education system to exploit the potential of institutional agreements as a complementary mechanism to strengthen accountability, transparency and focus on outcomes. As discussed above, carefully designed performance or quality agreements can be an effective way for institutions to demonstrate commitment to societally relevant goals and focus efforts to achieve these, while allowing differentiation and respecting the autonomy of institutions for designing and implementing their own strategies. Institutional agreements in other OECD jurisdictions that set out institutional strategies and planned actions in relation to shared goals have had positive effects on dialogue between public authorities and higher educations, strategic planning in institutions and communication and results-orientation among staff in the systems where they have been used.

Although models of institutional agreement vary, such systems generally have the advantage of allowing a more nuanced, qualitative approach to performance improvements and strategic development in a wide range of areas of activity than funding models driven by a limited set of standardised quantitative metrics. Moreover, agreements between institutions and government can more easily be designed as compacts between equal partners, relying on mutual trust, than models based on standardised indicators, with their inherent principal-agent relationship. Systems of institutional agreements require careful design. Key issues for higher education institutions and public authorities include how to specify goals against which progress can be measured, without encouraging a fixation on indicators, and how and when to assess achievement of the goals established. As discussed in this chapter, institutional agreements can be implemented without a direct link to institutional funding and, even in those systems where a link is made, the value of funding made conditional on achieving goals is typically limited. Nevertheless, evidence from OECD jurisdictions that have implement institutional agreements also suggests that such agreements work best when at least a modest level of additional funds is made available to incentivise and reward the engagement of institutions (de Boer et al., 2015[9]).

The lack of a recurrent funding allocation for strategic investments and the limited availability of funds for capital investment were identified as significant challenges for Flemish higher education institutions during the review interviews. The discussion in this chapter has highlighted that Flemish institutions are not alone and that few higher education funding models make explicit allocations for strategic investment. A notable exception to this is Finland, which allocates 15% of total public funding to universities and 5% of total funding for universities of applied science for strategic development. These funds are allocated to institutions in return for strategies negotiated in performance agreements. In light of the identified needs in the Flemish Community, there are likely to be further lessons to learn from the Finnish system.

The specific issue of funds for capital investment deserves specific attention. The level of annual public funding specifically for capital investment allocated in recent years (around EUR 33 million) is widely considered by Flemish stakeholders to be insufficient. At the same time, institutions are free to use core funding for capital investment, as many institutions do. Although there would appear to be a case for an increase in funding for capital investment, it is difficult to formulate specific recommendations on the level of resources needed for capital investment on the basis of the evidence available to this review. Particularly in light of competing pressures on public spending, further analysis of capital investment needs in the higher education sector and of options for financing these will be required at Flemish level. Given the comparatively low level of the funds earmarked for capital funding and the widespread practice of using other funds for capital investment at institutional level, the rationale for maintaining a separate funding line is questionable.

References

[6] Advisory Committee on Funding Higher Education and Research (2019), Wissels om Naar een transparante en evenwichtige bekostiging, en meer samenwerking in hoger onderwijs en onderzoek (Changing track; towards transparent and balanced funding and more cooperation in higher education and research), https://www.rijksoverheid.nl/documenten/rapporten/2019/05/01/bekostiging-hoger-onderwijs-en-onderzoek-wissels-om (accessed on 22 November 2019).

[40] Attewell, P., S. Heil and L. Reisel (2012), “What Is Academic Momentum? And Does It Matter?”, Educational Evaluation and Policy Analysis, Vol. 34/1, pp. 27-44, http://dx.doi.org/10.3102/0162373711421958.

[13] Audit Scotland (2019), Finances of Scottish universities, Auditor General for Scotland, Edinburgh, https://www.audit-scotland.gov.uk/report/finances-of-scottish-universities (accessed on 20 April 2021).

[34] Claeys-Kulik, A. and T. Estermann (2015), DEFINE Thematic Report: Performance-based Funding of Universities in Europe, European University Association, Brussels, https://eua.eu/downloads/publications/define%20thematic%20report%20performance-based%20funding%20of%20universities%20in%20europe.pdf (accessed on 24 October 2020).

[41] Clovis, M. and M. Chang (2019), “Effects of Academic Momentum on Degree Attainment for Students Beginning College at 2-Year Institutions”, Journal of College Student Retention: Research, Theory and Practice, http://dx.doi.org/10.1177/1521025119826245.

[21] Danish Government (2020), Finanslov for finansåret 2021 - 19 Uddannelses- og Forskningsministeriet (Finance law for the financial year 2021 - 19 Ministry of Education and Research), Danish Government, Copenhagen, https://fm.dk/media/18423/fl21a19.pdf (accessed on 28 June 2021).

[27] de Boer, H. and B. Jongbloed (2018), Evaluatie interne allocatiemodellen Vlaamse universiteiten (Evaluation of the internal allocation models in Flemish universities), CHEPS - Center for Higher Education Policy Studies, Universiteit Twente, https://data-onderwijs.vlaanderen.be/onderwijsonderzoek/project/1001 (accessed on 27 September 2021).

[9] de Boer, H. et al. (2015), Performance-based Funding and Performance Agreements in Fourteen Higher Education Systems: Report for the Ministry of Education, Culture and Science, Center for Higher Education Policy Studies (CHEPS), Enschede, https://research.utwente.nl/en/publications/performance-based-funding-and-performance-agreements-in-fourteen- (accessed on 4 January 2021).

[7] Deloitte Access Economics (2016), Cost of delivery of higher education - Final Report, Australian Government Department of Education and Training, https://docs.education.gov.au/system/files/doc/other/deloitte_access_economics_-_cost_of_delivery_of_higher_education_-_final_report.pdf (accessed on 29 January 2021).

[35] Dohmen, D. (2016), “Performance-based funding of universities in Germany: An empirical analysis”, Investigaciones de Economía de la Educación, Vol. 11, pp. 111-132, https://econpapers.repec.org/bookchap/aecieed11/11-05.htm (accessed on 4 February 2021).

[15] DUO (2021), WO Gegevensboek 2015-2019 (Universities Databook 2015-2019), Dienst Uitvoering Onderwijs, https://www.duo.nl/open_onderwijsdata/databestanden/onderwijs-algemeen/financiele-cijfers/gegevensboeken.jsp (accessed on 24 May 2021).

[32] Dutch Government (2021), Referentieraming OCW 2021 (Reference estimate by the Ministry of Education, Culture and Science, 2021), Dutch Government, https://www.rijksoverheid.nl/documenten/rapporten/2021/04/19/referentieraming-ocw-2021 (accessed on 31 May 2021).

[43] Dutch Government (n.d.), Kwaliteitsafspraken hoger onderwijs, https://www.rijksoverheid.nl/onderwerpen/hoger-onderwijs/kwaliteitsafspraken-hoger-onderwijs.

[3] Flemish Government (2021), Leerkrediet (Learning Credit), Flemish Department of Education and Training, https://onderwijs.vlaanderen.be/nl/leerkrediet (accessed on 28 May 2021).

[44] Flemish Government (2020), 2020-2024 MJ Raming (2020-2024 Multi-annual (budget) estimation), Flemish Government, Brussels, https://publicaties.vlaanderen.be/view-file/31866 (accessed on 27 September 2021).

[26] Flemish Government (2020), Conceptnota: onderwijsbelastingseenheden van de hogescholen: evaluatie en groeipad in meerjarig perspectief (Concept note: subject-area weightings in the university colleges: evaluation and growth path in a multi-annual perspective), https://publicaties.vlaanderen.be/download-file/38462 (accessed on 27 May 2021).

[19] Flemish Government (2020), Verslag over de financiële toestand en de evolutie van het personeelsbestand van het hoger onderwijs - Deel I: Hogescholen (Report on the financial situation and the evolution of staffing in higher education - Part 1 University Colleges), Regeringscommissariaat Hoger Onderwijs, Brussels, https://www.vlaanderen.be/publicaties/verslag-over-de-financiele-toestand-en-de-evolutie-van-het-personeelsbestand-van-het-hoger-onderwijs (accessed on 16 January 2021).

[14] Flemish Government (2020), Verslag over de financiële toestand en de evolutie van het personeelsbestand van het hoger onderwijs in 2019. Deel II - Universiteiten (Report on the financial situation and the evolution of staffing in higher education - Part 2 Universities), Regeringscommissariaat Hoger Onderwijs, Brussels, https://www.vlaanderen.be/publicaties/verslag-over-de-financiele-toestand-en-de-evolutie-van-het-personeelsbestand-van-het-hoger-onderwijs (accessed on 12 January 2021).

[1] Flemish Government (2015), Evaluatie van een aantal aspecten van het financieringsmechnisme in het hoger onderwijs (Evaluation of a number of aspects of the higher education funding mechanism), Flemish Government - Department of Education and Training, Brussels, https://www.onderwijs.vlaanderen.be/sites/default/files/atoms/files/Evaluatie-financiering-Hoger-Onderwijs.pdf (accessed on 4 May 2020).

[2] Flemish Government (2013), Codex Hoger Onderwijs (Higher Education Code), https://data-onderwijs.vlaanderen.be/edulex/document/14650 (accessed on 28 May 2021).

[39] Flemish Parliament (2019), Answer to a written question from Koen Daniëls. Studenten hoger onderwijs - Evolutie opgenomen studiepunten en studierendement (Students in higher education - evolution of enrolled credits and pass rates), Flemish Parliament, Brussels, https://docs.vlaamsparlement.be/ (accessed on 1 June 2021).

[5] Flemish Parliament (2018), Ontwerp van decreet betreffende de uitbouw van de graduaatsopleidingen binnen de hogescholen en de versterking van de lerarenopleidingen binnen de hogescholen en universiteiten (Draft decree concerning the expansion of associate degree programmes in the university colleges and the strengthening of teaching education programmes in the university colleges and universities), https://docs.vlaamsparlement.be/pfile?id=1379600 (accessed on 29 May 2021).

[42] Frontier Economics Ltd. (2015), A review of HEFCE capital expenditure, https://core.ac.uk/download/pdf/74378698.pdf (accessed on 30 May 2021).

[24] Golden, G., L. Troy and T. Weko (2021), “How are higher education systems resourced? Evidence from an OECD policy survey”, OECD Education Working Papers, No. 259, OECD Publishing, Paris, https://doi.org/10.1787/0ac1fbad-en.

[12] HEA (2019), Financial Trend Analysis - Universities and IoTs 2007-2017, Higher Education Authority, Dublin.

[33] HEA (2019), Performance Funding Allocations 2019, Higher Education Authority, https://hea.ie/assets/uploads/2017/04/Performance-Funding-Allocations-2019.pdf (accessed on 9 August 2021).

[29] Hemelt, S. et al. (2018), “Why is Math Cheaper than English: Understanding Cost Differences in Higher Education”, Discussion Paper Series, No. 11968, IZA Institute of Labor Economics, http://ftp.iza.org/dp11968.pdf (accessed on 27 September 2021).

[28] KPMG LLP (2019), Understanding costs of undergraduate provision in Higher Education, Government Social Research, London, https://www.gov.uk/government/publications/cost-of-undergraduate-higher-education-provision (accessed on 15 March 2021).

[36] O Shea, S. and J. O Hara (2020), “The impact of Ireland’s new higher education system performance framework on institutional planning towards the related policy objectives”, Higher Education, Vol. 80/2, pp. 335-351, http://dx.doi.org/10.1007/s10734-019-00482-5.

[30] OECD (2020), Labour Market Relevance and Outcomes of Higher Education in Four US States: Ohio, Texas, Virginia and Washington, Higher Education, OECD Publishing, Paris, https://dx.doi.org/10.1787/38361454-en.

[25] OECD (2020), Resourcing Higher Education: Challenges, Choices and Consequences, Higher Education, OECD Publishing, Paris, https://dx.doi.org/10.1787/735e1f44-en.

[8] OECD (2008), Tertiary Education for the Knowledge Society: Volume 1 and Volume 2, OECD Reviews of Tertiary Education, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264046535-en.

[10] Ortagus, J. et al. (2020), “Performance-Based Funding in American Higher Education: A Systematic Synthesis of the Intended and Unintended Consequences”, Educational Evaluation and Policy Analysis, Vol. 42/4, pp. 520-550, http://dx.doi.org/10.3102/0162373720953128.

[16] Rathenau Instituut (2020), Ontwikkeling derde geldstroom en beïnvloeding van wetenschappelijk onderzoek (Development of the third funding stream and its influence on scientific research), https://www.rathenau.nl/nl/vitale-kennisecosystemen/ontwikkeling-derde-geldstroom-en-beinvloeding-van-wetenschappelijk (accessed on 21 April 2021).

[37] Reviewcommissie Hoger Onderwijs en Onderzoek (2017), Prestatieafspraken: het vervolgproces na 2016 - Advies en zelfevaluatie (Performance Agreements: follow up after 2016 - Recommendations and self-evaluation, Reviewcommissie Hoger Onderwijs en Onderzoek, Den Haag, https://www.eerstekamer.nl/overig/20170321/rapport_review_commissie_met_het/document (accessed on 1 February 2021).

[38] Statistiek Vlaanderen (2021), Drop-out in het hoger onderwijs (Drop-out in higher education), Statistiek Vlaanderen, https://www.statistiekvlaanderen.be/nl/drop-out-in-het-hoger-onderwijs (accessed on 1 June 2021).

[20] UFM (2021), Regnskab — Uddannelses- og Forskningsministeriet (Financial Accounting - Ministry of Education and Research), Danish Ministry of Education and Research, https://ufm.dk/uddannelse/videregaende-uddannelse/professionshojskoler/okonomi/regnskab (accessed on 24 May 2021).

[31] UK Government (2020), Student number controls, UK Government, https://www.gov.uk/government/publications/student-number-controls (accessed on 31 May 2021).

[17] Universities Denmark (2021), Det Statistiske Beredskab - Danske Universiteter (University Statistics - Universities Denmark), https://dkuni.dk/tal-og-fakta/beredskab/ (accessed on 24 May 2021).

[23] Vereniging Hogescholen (2020), Financiën (Finances), Netherlands Association of Universities of Applied Sciences (Vereniging Hogescholen), https://www.vereniginghogescholen.nl/kennisbank/feiten-en-cijfers/artikelen/financien (accessed on 24 May 2021).

[22] Vipunen (2021), University income statement, Vipunen - Finnish Education Statistics, https://vipunen.fi/fi-fi/yliopisto/Sivut/Talous.aspx (accessed on 24 May 2021).

[18] Vipunen (2021), University of Applied Sciences Income Statement, Vipunen - Finnish Education Statistics, https://vipunen.fi/fi-fi/amk/Sivut/Talous.aspx (accessed on 24 May 2021).

[45] VSNU (2021), Kwaliteitsafspraken (Quality Agreements), Association of Universities Netherlands, https://www.vsnu.nl/kwaliteitsafspraken.html (accessed on 10 June 2021).

[4] Werkgroep "Studievoortgangbewaking" (2014), Werkgroep “Studievoortgangbewaking” - Eindrapportering (Working Group on Study Progress Monitoring - Final Report), Flemish Department of Education and Training, Brussels, https://onderwijs.vlaanderen.be/sites/default/files/atoms/files/Rapport-Studievoortgangsbewaking.pdf (accessed on 18 May 2021).

[11] Zacharewicz, T. et al. (2019), “Performance-based research funding in EU Member States—a comparative assessment”, Science and Public Policy, Vol. 46/1, pp. 105-115, http://dx.doi.org/10.1093/scipol/scy041.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2021

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.