This article was downloaded by: [University of Arizona]On: 04 July 2014, At: 22:37Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Click for updates

Regional StudiesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/cres20

‘Policy Scepticism’ and the Impact of Scottish HigherEducation Institutions (HEIs) on their Host Region:Accounting for Regional Budget Constraints underDevolutionKristinn Hermannssona, Katerina Lisenkovab, Peter G. McGregora & J. Kim Swalesa

a Fraser of Allander Institute, Department of Economics, University of Strathclyde, SirWilliam Duncan Building, 130 Rottenrow, Glasgow G4 0GE, UKb National Institute of Economic and Social Research (NIESR), 2 Dean Trench Street,Smith Square, London SW1P 3HE, UKPublished online: 12 Jul 2012.

To cite this article: Kristinn Hermannsson, Katerina Lisenkova, Peter G. McGregor & J. Kim Swales (2014) ‘PolicyScepticism’ and the Impact of Scottish Higher Education Institutions (HEIs) on their Host Region: Accounting for RegionalBudget Constraints under Devolution, Regional Studies, 48:2, 400-417, DOI: 10.1080/00343404.2012.697139

To link to this article: http://dx.doi.org/10.1080/00343404.2012.697139

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Policy Debates

‘Policy Scepticism’ and the Impact of ScottishHigher Education Institutions (HEIs) on theirHost Region: Accounting for Regional Budget

Constraints under Devolution

KRISTINN HERMANNSSON*, KATERINA LISENKOVA†, PETER G. MCGREGOR* andJ. KIM SWALES*

*Fraser of Allander Institute, Department of Economics, University of Strathclyde, Sir William Duncan Building,130 Rottenrow, Glasgow G4 0GE, UK. Emails: kristinn.hermannsson@strath.ac.uk, p.mcgregor@strath.ac.uk and

j.k.swales@strath.ac.uk†National Institute of Economic and Social Research (NIESR), 2 Dean Trench Street, Smith Square, London SW1P 3HE, UK.

Email: k.lisenkova@niesr.ac.uk

(Received July 2010: in revised form March 2012)

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. and SWALES J. K. ‘Policy scepticism’ and the impact of Scottish highereducation institutions (HEIs) on their host region: accounting for regional budget constraints under devolution, Regional Studies. A‘policy scepticism’ has emerged that challenges the results of conventional regional higher education institution (HEI) impact ana-lyses. This paper provides a systematic critique of such scepticism. While rejecting its extreme form, the limiting effect of thebinding public-sector expenditure constraints under devolution is noted and it is shown how conventional impact analyses canbe augmented to accommodate these constraints. While the results suggest that conventional impact studies overestimate theexpenditure impacts of HEIs, they also demonstrate that the policy scepticism that treats these expenditure effects as irrelevantneglects some key aspects of HEIs, in particular their export intensity.

Higher education institutions (HEIs) Input–output Scotland Impact study Multipliers Devolution

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. and SWALES J. K. “政策存疑论”与苏格兰高等教育机构（HEIs）对其所在区域之影响：阐述中央权力下放的区域预算限制，区域研究。“政策存疑论”的兴起，挑战了传统的区域高等教育机构影响分析之成果。本论文则对该存疑论提出了系统性的批判。本研究拒绝其极端形式，指出在中央权力下放之下，具有约束力的公部门支出限制仍有限制性的作用，亦证明如何得以强化传统的影响分析来应对这些限制。

研究结果虽指出传统的影响分析对高等教育机构支出的影响有着过高的评价，却也同时证明将这些支出效应视为无关紧要的政策存疑论，实则忽略了高等教育机构的某些重要面向，特别是它们的出口密集度。

高等教育机构（HEIs） 投入－产出 苏格兰 影响分析 乘数 中央权力下放

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. et SWALES J. K. Le ‘scepticisme envers la politique’ et l’impact des éta-blissements d’enseignement supérieur (EES) en Écosse sur la région d’accueil: tenir compte des contraintes budgétaires régionalesen vertu du transfert de compétences, Regional Studies. Du ‘scepticisme envers la politique’ a vu le jour, remettant en cause lesrésultats des analyses d’impact régionales conventionnelles des établissements d’enseignement supérieur (EES) en Écosse. Cetarticle fait la critique systématique d’un tel scepticisme. Tandis que l’on le rejette sous sa forme extrême, on note l’effet limitatifdes contraintes budgétaires engagées du secteur public suite au transfert de compétences et on démontre comment on pourraitcompléter les analyses d’impact conventionnelles pour tenir compte de ces contraintes. Alors que les résultats laissent supposerque les études d’impact conventionnelles surestiment l’influence des EES sur les dépenses, ils démontrent aussi que le scepticismeenvers la politique qui considère cette influence sur les dépenses comme dénuée d’intérêt, ne fait aucune attention à certaines car-actéristiques fondamentales des EES, notamment leur propension à l’exportation.

Regional Studies, 2014

Vol. 48, No. 2, 400–417, http://dx.doi.org/10.1080/00343404.2012.697139

© 2012 Regional Studies Associationhttp://www.regionalstudies.org

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

Établissements d’enseignement supérieur (EES) Entrées-Sorties Écosse Étude d’impact Multiplicateurs Décentralisation

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. und SWALES J. K. ‘Politische Skepsis’ und die Auswirkung derschottischen höheren Lehranstalten auf ihre Gastregion: Berücksichtigung von regionalen Etateinschränkungen im Rahmender Dezentralisierung, Regional Studies. Es ist eine ‘politische Skepsis’ entstanden, die die Ergebnisse der herkömmlichen Analysenüber die regionalen Auswirkungen von höheren Lehranstalten in Frage stellt. In diesem Beitrag wird diese Skepsis auf systematischeWeise kritisiert. Ihre extreme Form wird abgelehnt, doch zugleich weisen wir auf die einschränkende Auswirkung der verbind-lichen Ausgabengrenzen für den öffentlichen Sektor im Rahmen der Dezentralisierung hin und zeigen, wie sich die herkömm-lichen Auswirkungsanalysen erweitern lassen, um diese Einschränkungen zu berücksichtigen. Die Ergebnisse lassen daraufschließen, dass die Auswirkungen auf die Ausgaben höherer Lehranstalten in den herkömmlichen Auswirkungsanalysenüberschätzt werden, doch sie zeigen auch, dass die politische Skepsis, die diese Ausgabenauswirkungen als irrelevant behandelt,einige zentrale Aspekte der Lehranstalten vernachlässigt, insbesondere ihre Exportintensität.

Höhere Lehranstalten Input-Output Schottland Auswirkungsstudie Multiplikatoren Dezentralisierung

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. y SWALES J. K. ‘Escepticismo político’ y el impacto de las instituciones deeducación superior escocesas en su región anfitriona: considerar las restricciones del presupuesto regional en la transferencia decompetencias, Regional Studies. Ha surgido un ‘escepticismo político’ que cuestiona los resultados de los análisis convencionalesdel impacto regional de las instituciones de educación superior. En este articulo incluimos una crítica sistemática de tal escepticismo.A la vez que se rechaza su forma extrema, observamos el efecto limitado de las restricciones de los gastos del sector público en latransferencia de competencias y demostramos cómo pueden aumentarse los análisis convencionales de impacto para tener en cuentaestas restricciones. Si bien los resultados indican que los estudios convencionales de impacto sobreestiman los impactos de los gastosde las instituciones de educación superior, también demuestran que el escepticismo político que considera que estos efectos de losgastos son irrelevantes se olvida de algunos aspectos importantes de las instituciones de educación superior, en concreto su intensi-dad exportadora.

Instituciones de educación superior Entrada-Salida Escocia Estudio del impacto Multiplicadores Transferencia decompetencias

JEL classifications: H75, I23, R15, R51

INTRODUCTION

There have been numerous studies of the impact ofhigher education institutions (HEIs) on their hostregional economies that focus solely on their effect onthe local demand for goods and services (for reviews,see, for example, FLORAX, 1992; and MCGREGOR

et al., 2006). These demand-side studies treat a univer-sity like any other business which demands goods andfactor services within the region. The best of thesestudies employ regional input–output (IO) analysis,though a ‘policy scepticism’ has emerged that challengesthe value of such analyses. This scepticism asserts thateither binding demand-side budget constraints orsupply-side resource constraints ‘crowd out’ HEIexpenditure effects on the host regional economy, tothe point where the regional impact of HEIs’ expendi-tures is taken to be negligible. A further complementingfactor is the perception of a moral hazard in evaluationstudies, which undermines the credibility of impactstudies. The present provides a systematic critique ofthis view. However, while the extreme form of policyscepticism is rejected, the importance of bindingpublic sector budget constraints under UK devolutionis acknowledged, and it is argued that future regionalimpact studies should be modified, where appropriate,

to accommodate these constraints. Furthermore, theimportance of methodological rigour and transparencyof assumptions is acknowledged in order to maintainthe credibility of impact studies.

The argument is illustrated through an application toScotland. This is a United Kingdom region with a largehigher education sector and partially devolved fiscalresponsibilities. The choice of Scotland is particularlyappropriate for two reasons: Scotland’s devolved statusimposes a binding public sector expenditure constraintat the regional level; and the availability of relevantdata for the Scottish economy and Scottish HEIsallows a degree of confidence in the results that ismore difficult to replicate for other regions in theUK. However, it should be emphasized that thisapproach is generally applicable to all impact studies ofpublicly funded activities in regions with a devolvedbudget.

The analysis of HEI impacts is based upon a purpose-built, HEI-disaggregated set of IO accounts for Scot-land, in which the higher education sector is separatelyidentified.1 The expenditure impacts of HEIs werederived using standard IO assumptions. However, itwas also considered how these assumptions, and theirassociated current practice, have to be modified toaccommodate the binding budget constraint of the

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 401

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

Scottish Government. A novel treatment of studentexpenditure impacts was also implemented in which,in line with standard IO assumptions, the authorssought to identify the degree to which students’ con-sumption expenditures could be treated as exogenous.

This paper focuses exclusively on the expenditureimpacts of HEIs and their students. However, it mustbe stressed that these are only part of a broader interactionbetween HEIs, the economy and the wider community.Importantly, HEIs stimulate the supply side of their hostregional economies through activities such as: improvingskills in the labour force (BLUNDELL et al., 2005;BRADLEY and TAYLOR, 1996; CHECCHI, 2006;HARMON and WALKER, 2003; PSACHAROPOULOS

and PATRINOS, 2004), undertaking knowledgeexchange (ACS, 2009; ANSELIN et al., 1997; FISCHER

and VARGA, 2003; PARKER and ZILBERMAN, 1993;VARGA and SCHALK, 2004), and contributing to inno-vation (ANDERSSON et al., 2009; ANSELIN et al., 2000;JAFFE, 1989; LUNDVALL, 2008). Recent evidencesuggests human capital plays a key role in the causallink from HEIs to innovation (FAGGIAN andMCCANN, 2009; FAGGIAN et al., 2010). Furthermore,a persuasive case has been made that a more educatedpopulation results in long-term indirect benefits, suchas improved public health and lower crime rates(MCMAHON, 2004, 2009). Supply-side impacts arepotentially large relative to demand-side impacts(HERMANNSSON et al., 2010c) and merit a systematicstudy at the regional level.2 However, that does notmean expenditure impacts are yet fully understood. Thispaper seeks to clarify some of those demand-side issues.

The remainder of the paper is structured as follows.The second section outlines the approach taken by con-ventional (IO-based) regional HEI impact studies andsummarizes the results from using this method. Thethird section explores the basis of ‘policy scepticism’. Itis argued that complete supply-side crowding out isnot applicable in the context of a single devolvedregion. However, the recognition of a binding govern-ment expenditure constraint should be incorporatedinto HEI impact calculations in the case of UK devolvedregions such as Scotland. In so far as increased HEIactivity in Scotland is financed by the Scottish Govern-ment’s reducing of expenditure on other activities, thenthe effect of this expenditure switching should be expli-citly identified. The aggregate impact of expenditureswitching is much lower than the impact of a corre-sponding level of additional expenditure, though inthe Scottish case the net impact remains positive.

The fourth section shows that it would be wrong toinfer from the small net ‘balanced expenditure multi-plier’, which applies to Scottish general governmentexpenditure being switched to HEIs, that HEIs have anegligible net overall demand-side impact on theirhost region. A total of 45% of Scottish HEIs’ funds donot come from the Scottish Government, reflectingthe export intensity of Scottish HEIs as a group.

Brief conclusions are presented in the fifth section.Overall, the results suggest that conventional impactstudies overestimate the impact of HEIs’ expenditureson their host region. However, it is also demonstratedthat the policy scepticism that treats the expenditureeffects of HEIs as irrelevant neglects some importantcharacteristics of these institutions, notably their exportintensity. The measured impact of Scottish HEIs isused to illustrate this analysis. However, the approachis relevant to measuring the impact of activities requiringany degree of public funding in regions where budgetsare devolved.3

CONVENTIONAL REGIONAL IMPACTANALYSES

Conventional demand-side impact analysis of HEIs ontheir host regions identifies the total effect as the sumof the impact of institutional expenditures and of (typi-cally part of) the expenditures of their students. Thissection begins with a brief account of regional IO-based impact analyses, starting with institutional, andthen subsequently student, expenditures.

Theoretical basis of conventional regional impact analyses

Regional impact analysis is frequently employed tocapture the total spending effects of institutions, projectsor events. In addition to identifying the direct spendinginjection, multiplier, or ‘knock-on’, impacts are esti-mated by summing the subsequent internal feedbackswithin the economy. This section briefly outlines thestandard methods adopted by impact studies.4 It usesthese standard methods to derive the conventionalexpenditure, or demand-side, impacts of the HEIsector on the Scottish economy for 2006.

Most regional demand-driven models (for example,Export base, Keynesian multiplier, Input–Output)make a crucial distinction between exogenous andendogenous expenditures. In these models, exogenousexpenditures are those determined independently ofthe level of activity within the relevant economy,whilst endogenous variables are driven by, and thereforedependent upon, such activity. Specifically, the demandfor intermediate inputs, and often the domestic demandfor consumption goods, is taken to be endogenous.Other elements of final demand (exports, governmentexpenditure and investment) are typically taken to beexogenous.5 The models therefore trace a clear causalpathway from exogenous to endogenous expenditure.

In the standard Leontief Input–Output approach,total activity within the economy can be expressed asthe product of final demand and a multiplier process,represented by the Leontief inverse. This is summarizedas:

q = 1− A( )−1f (1)

402 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

where q is the vector of gross outputs; f is the vector offinal demands and (1 – A)−1 is the Leontief inversematrix. The output multiplier for any sector is thechange in total output for the economy as a wholeresulting from a unit change in the final demand forthat sector. It can be found as the sum of the entriesin the relevant column of the Leontief inverse. Thisallows a convenient expression for the gross output qiattributable to the final demands fi for the output ofsector i:

qi = lifi (2)

where li is the output multiplier for sector i.Although a number of variants are available, the

Type I and Type II demand-driven multipliers usedhere are typical for IO-based impact studies. Type Imultipliers incorporate the increase in demand for inter-mediate inputs but treat household consumption asexogenous. Type II multipliers include both intermedi-ate and household consumption demands as endogen-ous. Multipliers can be derived that relate a variety ofactivity outcomes to changes in exogenous demands.Examples are employment, income, output or grossdomestic product (GDP) multipliers (for furtherdetails, see HERMANNSSON et al., 2010a; and MILLER

and BLAIR, 2009, ch. 6).A key assumption underpinning the conventional

interpretation of these demand-driven models is thatthe supply side of the economy operates passively. Thatis to say, a change in demand in a particular sector gener-ates a proportionate change in supply, where theadditional output is supplied in exactly the same way,with exactly the same production processes, as the orig-inal output. At the regional level, multiplier analysis isconventionally validated by assuming one of two setsof conditions. In the short and medium run, passivesupply is justified by general excess capacity and regionalunemployment. In the long run, it is supported by factorsupplies effectively becoming infinitely elastic, asmigration and capital accumulation ultimately eliminateany short-run capacity constraints (MCGREGOR et al.,1996).6

The authors base the present study on an adjustedversion of the official Scottish Input–Output Table(HERMANNSSON et al., 2010a). Income and expendi-ture data for Scottish HEIs are used to specify a separateHEIs sector. That is to say, the education sector given inthe official IO table is split into two elements: HEI andnon-HEI education. This disaggregation necessarilyidentifies the income and expenditure structure of Scot-tish HEIs and makes it possible to derive appropriate(sector-specific) multipliers. The table and associatedmodel treat the higher education sector as any othersector: as a demander of goods and services and factorinputs, and as a supplier of services to meet intermediateand final demand. Applying these principles to derivethe demand-side impacts of HEIs entails estimating

the economic activity contingent upon the economy’sfinal demand for the HEIs’ services.

In addition to the impact of the institutions’ ownexpenditures, a further impact that needs to beaccounted for is the implicitly linked (exogenous) stu-dents’ consumption expenditure that occurs withinthe local economy. In practice this involves: determin-ing the level of student spending; judging the extentto which this is additional to the Scottish economy;and identifying how student expenditures are distribu-ted among sectors. The most difficult part of thisprocess is the disaggregation of students’ consumptionexpenditures into its exogenous and endogenouscomponents.

The regional impact of HEIs’ own expenditures

An extensive literature estimates the demand-sideimpact of HEI spending on their host regional econ-omies solely through these expenditure-relatedeffects. FLORAX (1992) listed over forty studies of theregional economic impact of HEI expenditure. Muchhas been published since, and MCGREGOR et al.(2006) summarized the methods and findings of themain UK studies. Table 1 reviews the major ScottishHEI impacts studies. Most UK studies, especially theearlier work, are based on Keynesian income–expendi-ture models (for example, ARMSTRONG, 1993; BATTU

et al., 1998; BLEANEY et al., 1992; BROWNRIGG,1973), whilst a smaller number use some variant ofIO modelling (for example, BLAKE and MCDOWELL,1967; KELLY et al., 2004; and, most recently, HER-

MANNSSON et al., 2010b).7 These studies differ in thetype of multiplier they report, the approach used toderive the multiplier values, and the geographicalscale over which the impact is measured. Unsurpris-ingly, therefore, the multiplier values generated differsomewhat and are in many cases not directlycomparable.8

A variety of multipliers can be derived to link a par-ticular exogenous change to changes in a number ofeconomic activity metrics. The output multipliersrelate changes in final demand to the change in grossoutput. Therefore, a Type II output multiplier of2.15, as found in MCNICOLL (1993), implies that1 unit increase in the final demand for the outputs ofStrathclyde University (Glasgow) leads to a Scotland-wide change in output of £2.15. The stated employ-ment multipliers show the economy-wide change inemployment caused by a 1 unit increase in directemployment. The household income multiplier usedby BLAKE and MCDOWELL (1967) is slightly unusual,but appropriate for their small borough application,where they related changes in the total output of theUniversity of St Andrews to changes in local householdincome. The GDP multipliers used by BROWNRIGG

(1973) linked exogenous changes in regional GDP tothe overall change in regional GDP.9

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 403

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

When a conventional IO analysis is applied to thepresent authors’HEI-disaggregated IO table for Scotland,it was found that in 2006 the Type I output multiplier forthe HEIs sector was 1.33 and the Type II multiplier was2.12. That is to say, each £1 of final demand for theoutput of HEIs should generate Scotland-wide outputamounting to £1.33 if indirect knock-on effects areincluded and £2.12 if the induced impacts on householdconsumption are counted as well.

The treatment of students’ consumption expenditures

In past impact studies student expenditures have beentreated in alternative ways: one incorporates only theexpenditures of in-coming students (for example,KELLY et al., 2004), the other includes all student expen-ditures irrespective of their origin (for example, HARRIS,1997). Here it is argued that each of these past treatmentsrepresents an approximation to an IO accountingapproach in which the crucial distinction is betweenthe exogenous and endogenous components of studentexpenditures. While it is true that the whole of externalstudents’ expenditures can be regarded as exogenous tothe host region, home students’ expenditures cannotlegitimately be treated as either wholly endogenous,which is what would be required to validate the firstapproach, or wholly exogenous, which would berequired to validate the second.

A distinction is then made between three types ofstudent. These are: students from outwith the UK;

students from the UK previously domiciled outwithScotland; and students from Scotland. These are ident-ified as foreign, RUK and Scottish students, respect-ively. The treatment of foreign students isstraightforward: their expenditures are taken to beunambiguously exogenous, as their incomes areassumed to be derived wholly from an external location.The appropriate treatment of their expenditure is similarto that of tourists.

From a Scottish perspective, RUK students are verysimilar to foreign students in that most of their incomeis sourced externally. However, it is assumed that a pro-portion of their expenditure is covered by wages earnedlocally. In an IO framework, this expenditure simplydisplaces the expenditure of other local workers whowould have taken these jobs; the corresponding expen-diture is not exogenous.

For Scottish students, the distinction between theirendogenous and exogenous consumption is less clearcut. To a large extent their income, and hence con-sumption, is endogenous to the local economy in thatit comes not only from wages earned from local indus-tries, but also from transfers from within local house-holds. For Scottish students this paper adoptssimplifying assumptions that are in line with the standardIO notion of exogeneity. The exogenous componentsof local students’ consumption expenditures areassumed to be expenditures financed from commercialcredit taken out during their years of study, studentloans, and education-related grants and bursaries.

Table 1. Overview of main Scottish higher education institution (HEI) impact studies

Subject of study Multiplier value Geographic boundary Source of multiplier value

St Andrews University (BLAKE andMCDOWELL, 1967)

1.45 (household income) St Andrews (population10000)

Local Input–Output Table (BLAKE andMCDOWELL, 1967)

Stirling University (BROWNRIGG,1973)

1.24–1.54 (GDP) Parts of Sterling andPerth (population96 000)

Keynesian multiplier (BROWN, 1967;GREIG, 1971)

Strathclyde, Stirling and St AndrewsUniversities (LOVE and MCNICOLL,1988)

1.34, 1.43, 1.36 (studentspending)

Scotland Keynesian multiplier (BROWNRIGG andGREIG, 1975; MCNICOLL, 1981)

Aberdeen, Dundee and StirlingUniversities (LOVE and MCNICOLL,1990)

2.18 (output), 1.75 (GDP),1.95 (employment)

Scotland Scottish Input–Output Tables 1979

Aberdeen University (BATTU et al.,1998)

1.46 (spending), 1.61(employment)

North East of Scotland Keynesian multiplier (GREIG, 1971;BROWNRIGG, 1971; MCGUIRE, 1983;HARRIS et al., 1987)

Strathclyde University (KELLY et al.,2004)

1.63 (output Type II), 1.38(employment, Type II)

Scotland Scottish Input–Output Tables 2000

Strathclyde University (MCNICOLL,1993)

2.15 (output Type II) Scotland Scottish Input–Output Tables 1989

Scottish HEIs (1) (MCNICOLL, 1995) 1.76 (output Type II), 1.7(employment Type II)

Scotland Scottish Input–Output Tables 1994

Scottish HEIs (2) (MCNICOLL et al.,1999)

1.73 (output Type II), 1.42(employment Type II)

Scotland Scottish Input–Output Tables 1997

Scottish HEIs (3) (MCNICOLL et al.,2003)

1.6 (output Type II), 1.4(employment Type II)

Scotland Scottish Input–Output Tables 2004

HEI impacts project (HERMANNSSON

et al., 2010a, 2010b)1.3 (output Type I), 2.1

(output Type II)Scotland Scottish Input–Output Tables 2004

Note: GDP, gross domestic product.

404 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

For information on Scottish students’ income andexpenditures this paper draws on a comprehensivesurvey by WARHURST et al. (2009). A more detailedaccount of the treatment of student expenditures isgiven in Appendix A. The multiplier impact of this con-sumption is calculated using a student expenditurevector estimated by KELLY et al. (2004). For studentconsumption spent on Scottish goods and services, theoutput multiplier is 1.8. However, the following adjust-ments need to be made to derive the impact on Scottisheconomic activity from student consumption. First,almost one-third of student consumption expenditureis on imports to Scotland. Second, the non-exogenousexpenditure needs to be removed from the impact cal-culation. After these adjustments the multiplier valuesfor foreign, RUK and Scottish student consumptionare 1.22, 0.89 and 0.50, respectively. This implies that£100 of consumption expenditure from foreign stu-dents in Scotland supports £122 of Scottish outputwhilst the corresponding expenditure from a localstudent supports £50 of Scottish output.

The aggregate output supported by the expenditures of HEIs andtheir students in Scotland

Table 2 shows the aggregate output supported by theexpenditures of Scottish HEIs, together with the corre-sponding figures for student consumption. This sectionbegins with the institutional expenditure. The directexpenditures of Scottish HEIs in 2006 are £1913million, measured in 2006 prices. This equals just over1% of total Scottish output. The indirect and inducedType II multiplier effects are given as £2148 million, sothat the conventional, demand-driven impact calculationascribes 2.28% of Scottish output, either directly orindirectly to theHEI sector.The associated direct employ-ment is 31520, whichmakes up nearly 1.6% of total Scot-tish employment. When multiplier effects are added thelevel of employment conventionally taken to be sup-ported by the HEI sector is 55136, just over 2.75% ofScottish total employment. The greater impact of HEIexpenditure on employment derives from the fact thatthe sector itself is relatively labour intensive so that a rela-tively large share of the initial expenditure is on labour.

The demand-side impacts associated with studentexpenditure quantify the output and employmenteffect of the additional exogenous consumption expen-diture made by students. The direct output is the dom-estic output purchased by this exogenous expenditure.It amounts to £394 million, which is just over 0.2% ofScottish output. Again multiplier effects generate anadditional £314 million, implying that Scottishstudent exogenous expenditure makes up, directly orindirectly, 0.4% of Scottish output. The associateddirect employment is calculated as the direct employ-ment required to produce the direct output. This is3545 full-time-equivalent (FTE) jobs with an associ-ated 3334 indirect and induced FTEs generating atotal of 6879. This is 0.34% of total Scottishemployment.

It was noted in the second section that the impact ofa given amount of expenditure by students fromoutwith the UK or by students from the rest of theUK has a bigger demand impact on the Scottisheconomy because a bigger proportion of the expendi-ture is exogenous. It is also the case that expenditureper head is higher for foreign students. However,despite the relatively modest per student impact, Scot-tish students make up 70% of the student populationand therefore drive approximately 50% of the totalstudent consumption impact. The significance of theconsumption spending of students from the rest of theworld is little less at approximately 30% of the totalconsumption impact, whereas the remaining 20% ismade up by the expenditure of students from the restof the UK.

This section has outlined best practice for estimatingthe demand-driven impact of HEIs and their associatedstudent population using the conventional IO approach.That is to say, it has attempted to identify exogenousfinal demand expenditures and then calculate the indir-ect and induced effects using a standard IO multiplier.The results suggest that in 2006 the expendituresmade by Scottish HEIs and their students directly orindirectly supported almost 2.7% of Scottish outputand 3.1% of Scottish employment. The next sectionexamines the policy scepticism aimed at this approachand considers appropriate responses.

Table 2. Summary of expenditure impacts of higher education institutions (HEIs), based on traditionalinput–output (IO) assumptions

Final demand fordomestic output

Indirect and inducedimpacts Total impact

HEIs’ output (£, millions) 1913 1.07% 2148 1.20% 4062 2.28%HEIs’ employment (FTEs) 31 520 1.58% 23 616 1.18% 55 136 2.76%Students’ output (£, millions) 394 0.22% 314 0.18% 708 0.40%Students’ employment (FTEs) 3545 0.18% 3 334 0.17% 6879 0.34%

Total output (£, millions) 2308 1.29% 2 463 1.38% 4770 2.67%Total employment (FTEs) 35 065 1.76% 26 949 1.35% 62 014 3.10%

Note: Output is given as £ (millions); employment is given as full-time equivalents (FTEs) or as a percentage of the Scottish total.

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 405

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

POLICY SCEPTICISM AND THE IMPACTOF HEIs

There is a distinct degree of ‘policy scepticism’ aimed atmultiplier-based studies of the expenditure impacts ofHEIs. This has been expressed in private conversationswithin HEI policy circles in Scotland and the UK, aswell as in questioning during knowledge exchangeseminars and other public engagement activities, whereexpenditure impacts have been presented.10 The authorsdetect three sources for this scepticism. These are: thatdemand-driven multiplier studies do not recognizeresource restrictions; that they are often commissionedfor advocacy purposes and tend to overstate the potentialdemand-side knock-on effects; and that they fail to incor-porate public sector budget constraints. The cumulativeimpact of these criticisms is to undermine the credibilityof such studies within the policy community. However,it is argued that only a third of these concerns hasfundamental validity when considering the regionaldemand-side impacts of specific publicly funded activity.Further, in the case of HEIs, this reduces, but does noteliminate, their significant demand-driven impacts.

Resource constraint

One potentially important source of scepticism withinthe UK about regional demand-driven impact multi-pliers is the 100% crowding-out argument that charac-terizes the HM Treasury Green Book’s analysis ofregional impacts (HM TREASURY, 2003). Here a puredemand disturbance that stimulates employment inone region has an equal and offsetting impact onemployment in other regions of the UK, given thatthe UK economy is taken typically to operate at ‘fullemployment’ (or the natural rate of unemployment orNAIRU).

However, even if there were 100% crowding out atthe level of the UK as a whole, this would not apply atthe level of the host regional economy. This resultrequires fixed, fully employed resources. This is muchless credible at the regional level than at the nationallevel. In the short run regional policy is typicallyattempting to increase activity in an underemployedlocal economy, and in the long run both capital andlabour are mobile between regions. It is thereforequite legitimate for Scottish and Welsh governments,for example, to be concerned about the demand-sideimpact of particular institutions/expenditures for theirown economies. In this context, aggregate host–region employment multipliers are clearly not con-strained to be zero.

Moral hazard in evaluation studies

Demand-driven impact studies often adopt extremeassumptions in order to maximize the estimated econ-omic contribution of a particular institution or activity.

This line of criticism, as applied specifically to identify-ing the impact of HEIs, is picked up by SIEGFRIED et al.(2007), who concluded:

If these economic impact studies were conducted at thelevel of accuracy most institutions require of facultyresearch, their claims of local economic benefits wouldnot be so preposterous, and, as a result, trust in andrespect for higher education officials would be enhanced.

(p. 546)

For a similar criticism of demand-driven local impactstudies of biofuel initiatives, see SWENSON (2006).

The policy scepticism that this engenders is a classiccase of moral hazard in information transmission(OSBORNE, 2004, ch. 10). Those commissioning theimpact analysis have an interest in finding large effects.Further, the method of calculation, whilst conceptuallyrather straightforward, is technically quite complex.Therefore, policy-makers distrust the figures and oneresponse is simply to discount them altogether. Thepresent authors view this as a serious issue, but toignore possible large variations in demand-side effectswhen taking policy decisions is not sensible. The mostappropriate response is to require greater methodologi-cal rigour, which it is hoped will be shown in this paper.

Expenditure impacts under a budget constraint

The idea here is that an increase in public expenditureon HEIs will induce offsetting changes in demandthrough the operation of a binding regional publicsector expenditure constraint. In the UK, the Barnettformula allocates a block grant to each of the devolvedadministrations, a grant that is calculated on a populationbasis.11 The devolved administrations have limitedadditional sources of revenue and cannot borrow orcarry significant funds from one year to the next.12

Therefore, essentially UK devolved authorities, includ-ing the Scottish Government, have fixed annualbudgets. However, there are no constraints on the allo-cation of these budgets across the devolved policy areas,one of which is higher education. There is no ring-fencing of expenditure within the budget, so thatwhen the Scottish Government allocates expenditureto one activity, there is a clear opportunity cost interms of possible expenditures on other activities.Within this framework, variation in any specific formof public activity can be seen simply as expenditureswitching. From this perspective, policy-makers couldregard with extreme scepticism any demand-drivenargument for a particular type of public expenditure.This is because it replaces other public expenditureswhich would have broadly the same demand-sideimpact. The multiplier would then be approximatelyzero.

Given this context, it can be misleading for an impactstudy to treat the Scottish Government’s funding ofHEIs as an exogenous stimulus to the regional

406 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

economy, although that is standard IO practice. Theauthors regard this argument as valid and pursue theimplications of taking the government’s budget con-straint into account in identifying the demand-drivenimpact of HEIs.

EXPENDITURE SWITCHING, HEIs ANDPUBLIC BUDGET CONSTRAINTS

The authors wish to illustrate the importance of recog-nizing the public sector budget constraint and the con-sequent expenditure switching that this implies. Twosimulations are therefore conducted. In each a hypothe-tical additional £100 million of expenditure on HEIs inScotland is introduced. The first case adopts the tra-ditional impact study assumption that the exogenousincrease in expenditure is entirely externally funded,for example from UK-level funding or foreign students’fees. This would have no direct ramifications for otherpublic spending in Scotland. The second case examineshow the impacts change when there is a correspondingreduction of other public spending in Scotland. In thelatter case the offsetting £100 million reduction inpublic spending is applied to an aggregation of thosesectors that receive 93%13 of central and local govern-ment final demand in the Scottish IO tables.

The conventional Type II multiplier for the HEIssector is 2.12. Without any offsetting cutbacks in publicspending the additional spending on HEIs therefore hasan output impact of £212 million. Approximately halfof that impact is realized as a direct consequence ofincreased activity in the HEIs themselves, whereas theother half is generated via ‘knock-on’ effects in other

sectors, particularly the retail and service sectors. Thetotal change in output and employment, and the distri-bution across sectors, is summarized in Table 3. Theseimpacts are shown graphically in the darker shaded barsin Figs 1 and 2.

A more complex picture emerges with expenditureswitching. The Type II multiplier for other publicexpenditure in Scotland is 1.97. If an increase in HEIs’funding is met by cutbacks in other Scottish publicexpenditure, the ‘multiplier’ for switching is equal to2.12 – 1.97 = 0.15.14 That is to say, for every £100million directed from the public sector to HEIs, theoutput impact of switching is £15 million. In particular,the estimated import propensity of HEIs (13%) is lowerthan the public sectors’ import propensity (17%). There-fore, for every £1 spent on HEIs, more is retainedwithin the regional economy than for governmentspending in general. A qualitatively similar resultemerges in results for employment impacts.

The recognition of the regional budget constraintimplies that multiplier effects on individual sectors areno longer universally positive, as in the conventionalcase. The net changes are again shown in Table 3 andin the lighter shaded bars in Figs 1 and 2. In particular,there is a significant contraction in the public sectorand a net contraction in other sectors that are more sen-sitive to changes in general public expenditure ratherthan the expenditure on output in the HEI sector.‘Banking and financial services’ and the ‘Transport,post and communications’ sectors show small netreductions in activity. In a UK devolved context,changes in public expenditure, determined by theregional government and therefore financed throughBarnett, typically involve expenditure switching (and

Table 3. Impact of a £100 million increase in final demand for Scottish higher education institutions (HEIs)

Sector

Without expenditure switching With expenditure switching

Change in finaldemand

(£, millions)Output impact(£, millions)

Employmentimpact (FTE)

Change in finaldemand

(£, millions)Output impact(£, millions)

Employmentimpact (FTE)

Primary and utilities 0 8 37 0 2 9Manufacturing 0 14 99 0 7 50Construction 0 8 78 0 3 31Distribution and retail 0 21 364 0 3 61Hotels, catering, pubs, etc. 0 5 140 0 1 16Transport, post and

communications0 10 90 0 0 −4

Banking and financialservices

0 8 48 0 −2 −12

House letting and real estateservices

0 17 44 0 4 11

Business services 0 8 134 0 0 −6Public sector 0 6 95 −100 −103 −1735HEIs 100 101 1666 100 101 1662Other services 0 6 88 0 0 2

100 212 2882 0 15 86

Note: FTE, full-time equivalent.

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 407

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

certainly have an opportunity cost in terms of alternativeuses within the region), and the multiplier effects areaccordingly more subdued. Indeed, even the directionof the net impact cannot be known a priori. This is acrucial result that appears not to be widely appreciatedin existing impact studies.

As can be seen from the above analysis, care must betaken in determining the source of financing for anyimpact study applied to a region with a devolved budget.While the example of HEIs is used here, the principle isquite general. Devolution matters a great deal for theappropriate conduct of regional impact analyses.

Such expenditure switching might be interpreted asimplying that the impact of HEIs’ spending is verylimited at the Scottish level. However, while HEIs areoften thought to be part of the public sector, they arein fact non-profit organizations. An analysis of theirincome based on data from the Higher Education Stat-istics Agency (HESA) (HERMANNSSON et al., 2010a)reveals that just 55% of their income can be tracedback to the Scottish Government. Some 29% comesfrom sources outside Scotland and approximately 16%originates from households, businesses, charities andother institutions whose funding is independent of the

Fig. 2. Employment impact of a £100 million increase in final demand for Scottish higher education institutions (HEIs)(£, millions)

Fig. 1. Output impact of a £100 million increase in final demand for Scottish higher education institutions (HEIs) (£, millions)

408 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

Scottish Government. The external income is unam-biguously additional to the Scottish economy and it isreasonable to assume the latter part is as well. Even ifthe regional public sector budget constraint impliescomplete crowding out of public spending on HEIswithin the region, only a part of HEIs’ activities is pub-licly funded. In fact, HEIs are characterized by signifi-cant exports (to the rest of the UK and the rest of theworld), and changes in export demand do not triggerany offsetting expenditure switching among finaldemands. The sources of income of Scottish HEIs aresummarized in Fig. 3.15 The next section explores thesignificance of this pattern of funding for the attributionof HEI impacts on the host region.

ACCOUNTING FOR THE REGIONALBUDGET CONSTRAINT WITHIN THEINPUT–OUTPUT (IO) FRAMEWORK

The IO tables provide a useful accounting framework.Based on the distinction between exogenous (finaldemand) and endogenous (‘knock-on’) expenditure,all activity within the regional economy can be attribu-ted to elements of regional final demand. As discussed inthe previous section, there is a criticism levelled againstthe conventional way of deriving the economy-wideexpenditure impact of HEIs. This is that, given theirfunding arrangements in Scotland, attributing to HEIsthe impact of spending Scottish Government funds is

disingenuous. Such an impact is not so much causedby the HEIs per se as it is by the availability of fundsfrom the Scottish Government and potentially similarresults could be obtained if the funds were to beswitched to other public services. The IO framework,combined with detailed information about the incomesources of HEIs, enables a disaggregation of thesector’s impacts in terms of the origin of the exogenousfinal demands. This allows an analysis of the extent towhich the impacts attributed to the HEIs sector undera traditional IO approach should in fact be attributedto the expenditure of the Scottish Government.

In order explicitly to take account of the publicexpenditure switching effects, as discussed in thethird section, the impacts of the Scottish Government(‘Barnett’) funding are deducted from the overallexpenditure impact. The direct expenditure on theoutput of Scottish HEIs is divided into Barnettfunding (BF), which comes through the Scottish Gov-ernment, and other funding (OF), which includes allother sources, including exports to the rest of theUK and the rest of the World. The conventional attri-bution to HEIs, as performed in the second section, issimply:

(BF +OF)MH

where MH is the multiplier value for the HEIs sector.The results of this attribution are summarized inFig. 4. The adjusted attribution subtracts the

Fig. 3. Income structure of the higher education institutions (HEIs) sector in the HEI-disaggregated input–output (IO) tables

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 409

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

Barnett-funded element and its own multiplier effects,which is:

BF ∗ MP

where MP is the multiplier for the aggregated publicsector. The adjusted attrribution is therefore givenby equation (3):

(BF +OF)MH − BF ∗ MP

= OF ∗ MH + BF(MH −MP) (3)

whereMH–MP is the switchingmultiplier value identifiedin the third section. To summarize, the demand impact ofHEIs net of Scottish Government funding equals theimpact attributable to other funding sources OF *MH inaddition to the switching impact BF(MH –MP).

To clarify, the impact of Scottish Governmentfunding upon HEIs can be rearranged into a ‘generic’public expenditure impact and a ‘constrained’ impact.The demand impacts of the HEIs sector are illustratedin these terms in the lower bar of Fig. 4. As thediagram reveals, when the expenditure impact of HEIsis disaggregated according to the source of income,just under half of it can be classified as a generic publicsector impact, leaving just over half of it as a constrainedimpact. This constrained impact comprises the direct,indirect and induced effects of the HEI expenditurethat is not subject to the budget constraint of theBarnett funding received by the Scottish Governmentplus the switching impacts.

An exactly analogous argument can be made inrespect of the appropriate attribution of student

expenditure impacts. In this case one has:

(BFS +OFS)MS − BFS ∗ MP

= OFS ∗ MS + BFS(MS −MP) (4)

where BFS is students’ consumption final demandattributable to Scottish Government student support;16

OFS is students’ exogenous final demand for consump-tion from other sources; MS is the output multiplier forstudents’ consumption expenditures; and MP is theoutput multiplier for the public sector.

When students’ consumption expenditures are analysedin this way, the results are qualitatively different from thosefor the HEIs’ institutional expenditures. Primarily as aresult of its strong direct import component, the outputmultiplier associated with students’ consumption expendi-ture is smaller than that for public sector expenditure perse. In this case the Scottish Government gets a smallerdemand stimulus from expenditures on student supportthan from general public expenditure. In this case, theswitching multiplier is negative, whereas it is positive forHEIs’ institutional expenditures. Table 4 shows theimpact of students’ consumption expenditures togetherwith that of the Scottish HEIs institutional expenditure.

This section examines the impact attributable to theHEI sector in Scotland in more detail than is typical forimpact studies. The analysis reveals that there is some jus-tification for a degree of policy scepticism based on thebinding regional public budget constraint. Slightly lessthan half of the impact of the HEI sector in Scotland isa ‘generic’ public spending impact thatwould havemate-rialized anyway had the public funds been used to expand

Fig. 4. Output impact of higher education institutions (HEIs) disaggregated by origin of final demand. The upper bar shows theconventional components of the gross impact, while the lower bar breaks the impact into a generic public sector impact and constrained

impact by implementing expenditure switching (£, millions)

410 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

the host region’s public sector, although there is a smallpositive switching impact of public funding for HEIs’own expenditures, and a small negative switchingimpact for students’ consumption expenditures.However, the work also shows that the extreme formof policy scepticism, which argues that once the publicbudget constraint has been accounted for the impact ofthe HEIs’ expenditures on the host region is negligible,is not supported by the evidence. This result is primarilyattributable to the funding for HEIs and studentconsumption expenditures that comes from sourcesindependent of the Scottish Government.

CONCLUSIONS

A ‘policy scepticism’ appears to have developed that con-stitutes a major challenge to studies of the regional impactsof HEIs. In the limit this policy scepticism suggests that theexpenditure impacts ofHEIs on their host regions are neg-ligible, andcan therefore be ignored.Thebinding resourceconstraint rationale for policy scepticism is rejectedhereona priori grounds, but the significance of the bindingregional public sector budget constraint under devolutionis acknowledged. This constraint is built into an augmen-ted IO analysis using the authors’ purpose-built HEI-dis-aggregated IO table for Scotland. The results offer somesupport for policy scepticism in that it is estimated thathalf of the regional expenditure impacts of Scottish HEIsare attributable to public funding that could generatesimilar (though not identical) effects if put to alternativeuses, such as expansion of the public sector within thehost region. Conventional multiplier/impact analysesthereforedooverstate theexpenditure impacts attributableto HEIs per se. However, the remaining impact is none-theless substantial in the Scottish case, and certainly notnegligible, as the extreme form of policy scepticismimplies. In fact, if funds useddirectly tofinance theScottishpublic sector were instead used to finance HEIs, therewould be a small net positive multiplier effect reflectingthe lower import propensity of HEIs. But for similar

reasons the switching of public funds to students andaway from the public sector would have a net negativemultiplier impact.

Although the hypothetical notion of switchingpublic funding illustrates the point that expenditureimpacts vary between expenditure categories, as a prac-tical proposition the authors do not envisage policy-makers engaging in significant alterations of spendingpriorities based on expenditure impacts alone. Alteringpublic expenditure significantly is likely to be difficult,at least in the short run, and policy-makers shouldhave a wider agenda than just expenditure impacts.This paper has presented an attribution analysis basedon the existing state of affairs. However, it is impossibleto tell how the HEIs sector would respond to changes infunding at the margin. It is unclear if public funding is acomplement or substitute for other sources of funding.Therefore, the authors caution against a mechanicaluse of the approach to project the likely impacts of gov-ernment expenditure cuts.

The analysis is capable of extension in a number ofdirections. Firstly, it can be applied to individual HEIsas well as to the HEI sector as a whole. HERMANNSSON

et al. (2010b) show that there is considerable heterogen-eity among Scottish HEIs in terms of their dependenceon public funding, and they identified the significanceof this for the scale of ‘balanced expenditure’multipliers.Secondly, although the focus here is on the expenditureimpacts of HEIs, the principles apply equally to anysector of interest which is at least partly publiclyfunded. Naturally, the authors’ judgement aboutpolicy scepticism does not necessarily generalize: thiswill depend on the characteristics of both the sectorunder consideration and the region. Thirdly, the analysiscan clearly be applied, and indeed should be applied, toall impact analyses that involve any element of localpublic funding conducted for any region that issubject to a binding public expenditure constraint,most obviously Wales and Northern Ireland in theUK context. In these circumstances, researchers

Table 4. Summary of overall spending impacts attributable to higher education institutions (HEIs) by origin of final demand andtype of impact

Generic public sector impact Constrained impact Conventional impact

Institutional spending 1125 788 1913Knock on impacts 1091 885 1976Switching impact 172 172Institutional impact total 2216 1846 4062… as a percentage of total impact 55 45 100

Exogenous student spending 87 495 582Knock on impacts of students’ consumption 84 108 192Switching impact −65 −65Student’s consumption impact total 171 538 708… as a percentage of total impact 24 76 100

Total impact attributable to HEIs 2387 2384 4770… as a percentage of total impact 50 50 100

Note: Output is given as £ (millions).

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 411

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

seeking to identify the economic activity attributable toa particular sector should acknowledge the devolvedbudget constraint explicitly and identify the fraction ofactivity attributable to the public funds. In general thiswill reveal that a significant part of HEIs’ impact is infact a ‘generic’ public expenditure impact and in thelimit this may reveal the demand-side impact of particu-lar regional institutions to be effectively zero once theregional public budget constraint has been taken intoaccount. However, in the case of Scottish HEIs con-sidered in this paper, substantial impacts can be attribu-ted to HEIs’ activity, in addition to those driven entirelyby local public expenditures. Fourthly, the analysis mayalso be usefully applied to regions that are not subject toa binding expenditure constraint, such as the Englishregions in the UK context. Even where there is nobinding constraint on public expenditure at the (rel-evant) regional level, it may still be of interest to assessthe demand-side opportunity cost of the publicfunding involved by exploring the impact of theiralternative use within the region.

Finally, it should be noted that the analysis described inthis paper is, in common with conventional regionalimpact analyses, focused solely on the expenditure ordemand-side effects of HEIs. This is a rather restrictivecontext in which to consider policy impacts. Therefore,the authors would not, for example, advocate the use ofestimated net ‘balanced expenditure’multipliers to prior-itize public expenditures. In the case of HEIs the messagewould in any case be mixed: HEIs’ own institutionalexpenditures have a rather higher multiplier than publicexpenditure per se, but the reverse is true of students’expenditures funded by government grant. However,much more importantly in the case of HEIs, at least, isthat one would expect many of their impacts on regionaleconomies to come through the direct stimulation of thesupply side, for example, through their impact on the skillsof the host region’s labour force and through knowledgeexchange activities. These impacts can only be exploredin a framework that explicitly accommodates thesesupply-side effects, so that IO analyses are inadequate tothe task, even if, as here, they are augmented to accom-modate regional public expenditure constraints. Thismay be particularly important for policy given thatthere is some evidence that the supply-side impacts ofScottish HEIs may be large relative to their expenditureimpacts (for example, HERMANNSSON et al., 2010c).

A poignant example of the dilemma policy-makersface in balancing demand- and supply-side impacts isthe case of local and external students. As this paperreveals, there are significant expenditure impacts fromattracting external funding and incoming students.However, it is an open question whether emphasizingthe export role of HEIs is ultimately desirable for theScottish economy. If a focus on external income comp-lements the HEIs’ capacity for building human capitaland engaging in knowledge exchange it is clearly agood thing overall. However, if there is some trade-

off between focusing on external competitiveness ofthe institutions and their role in producing graduatesfor the local labour market and engaging with localbusinesses, the outcome would be ambiguous. Currentunderstanding of these potential trade-offs is verylimited but the authors hope that highlighting the gapwill encourage future research.

Acknowledgements – This paper is an output of the‘Overall Impact of Higher Education Institutions on RegionalEconomies’ project funded by the Economic and SocialResearch Council (ESRC), the Scottish Funding Council(SFC), the Higher Education Funding Council of England(HEFCE), the Higher Education Funding Council of Wales(HEFCW), and the Department for Employment and Learn-ing Northern Ireland (DELNI) (Grant Number RES-171-25-0032). The authors also acknowledge additional fundingfrom the Centre for Public Policy for Regions. They areindebted to Nikos Pappas for excellent research assistanceand to Ursula Kelly for helpful discussions and advice. Fur-thermore, the authors are grateful to the participants at ‘Scot-land’s Universities and the Economy: Impact, Value andChallenges’, 30 June 2009, in particular Ron Botham. Thisproject is one of nine funded through the Impact of HigherEducation Institutions on Regional Economies Initiative.

APPENDIX A

Appendix A presents the details of the authors’derivation of the impact of students’ consumptionexpenditures. It draws on a comprehensive surveyby WARHURST et al. (2009),17 who conducted alarge-scale survey complemented with face-to-faceinterviews. Warhurst et al. interviewed 1000 Scottishdomiciled undergraduate students at Scottish institutionsand estimated their average term-time expenditure at£6404 in the academic year 2007/2008. However,these results only refer to a part of students at ScottishHEIs as one-third come from outwith Scotland(HERMANNSSON et al., 2010c, tab. 5) and 19%(HESA, 2007, ‘Students in Higher Education’, tab. 0b)are postgraduates. Surveys have not been carried outrelating to the expenditure of students of RUK andROWorigin nor for Scottish domiciled postgraduate stu-dents. These students’ expenditures are expected to begreater as expenditures generally increase with age andthe year of study, and these students are staying awayfrom home and so must pay for accommodation in full.

According toWARHURST et al. (2009) Scottish dom-iciled undergraduates living independently spent onaverage £7187 in 2007/2008, while those living withtheir parents spent £5317. The expenditure level ofScottish students who are living independently is usedas a proxy for expenditures of incoming students.However, it is reasonable to expect incoming studentsto incur more costs than locals if only due to their unfa-miliarity with local conditions and their inability to drawon a social network, in contrast to local students. A higher

412 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

estimate for living costs is, for example, suggested by theIcelandic Student Loan Fund (LÍN), which estimatesstudent expenditures (apart from tuition fees) for an aca-demic year in Scotland at £8520.18 Here the rather con-servative approach is adopted that the average for Scottishdomiciled undergraduates is applied to all Scottish dom-iciled higher education students and the average expen-ditures of Scottish domiciled undergraduates livingindependently are applied to all incoming students.

A number of adjustments have to be applied to the‘gross’ student spending as reported by Warhurst et al.to conform with IO assumptions (for an outline ofWarhurst et al.’s main findings on student spending inScotland, see Table A2). In particular care must betaken to deduct non-additional (‘endogenous’) spend-ing components to avoid double counting. For Scottishdomiciled students this means that the components ofconsumption that are treated as additional (exogenous)are those that are attributable to student loans, the com-mercial credit that students take out to support them-selves, and student support and grants, as reported byWarhurst et al. This changes slightly when the budgetconstraint of public expenditures in Scotland isacknowledged as student support and grants are to a sig-nificant extent19 funded by the Scottish block grant andtherefore represent a reallocation of Scottish Govern-ment spending within Scotland (for a general discussion,see the third section). The student loans received byScottish students are, however, treated as additional asthey are provided by the Student Loans Company, aUK-level non-departmental public body. Informaltransfers within the family do not constitute additionalspending in Scotland as they are a reallocation of totalhousehold spending.20 Term-time labour market earn-ings are equally not additional to the Scottisheconomy as, under the IO assumption of a passivesupply-side, if the student were not earning that wageincome some other Scotland resident would be. Thatleaves other income, which is assumed to be endogen-ous to the Scottish economy21 and the students’income shortfall (expenditure in excess of income).Precise information is not available on the compositionof this income shortfall, but it is expected to constitutesome combination of informal income/credit not pre-viously accounted for and commercial credit. Newcommercial credit taken out by Scottish domiciled stu-dents represents an exogenous impact on the localeconomy, while informal credits are assumed to beobtained locally and therefore represent a transferwithin the economy rather than an additional impact.

Warhurst et al. provide information on the amount ofcommercial credit taken out by Scottish students duringtheir time of study, which is used to estimate the mag-nitude of this impact. Care must be taken to countonly the net commercial credits obtained as studentsrun up commercial debts during term time but typicallyrepay these to some extent between years. WARHURST

et al. (2009, tab. 4.15, p. 100) reported the amount of

commercial credit owed by students at the end ofeach of their year of study. They found a wide rangeof commercial debt incurred by year of study. Ofcourse it must be born in mind that Warhurst et al.’ssurvey is a cross-section, but interpreted literally itsuggests that students rely less on commercial credit asthey progress through their studies (and a net repaymentoccurs between years 3 and 4). This is in line with theirfindings that students’ earning power increases with yearof study. Here the assumption is adopted that commer-cial debt levels at the end of year 4 are representative oftheir overall net incurrence for the entire duration ofundergraduate study (Table A1).

Based on these assumptions the average additional(‘exogenous’) component of Scottish students’ term-timespending is £346 (£1384/4). The assumption suffersfromapotential downwardbias in that fourth-year studentsare less than one-quarter of the student population.However, it could be counter-argued that students willuse income earned in the following summer to makeadditional payments to their commercial debt. Availableevidence unfortunately does not allow a precise estimate,but on balance the assumption adopted here should be

Table A1. Commercial credit at the end of term time byyear (£)

Year1

Year2

Year3

Year4

Commercial credit owed at theend of term time

968 1240 1699 1384

Net change in commercial debtbetween years of study

968 272 459 −315

Implied average per year of study 968 620 566 346

Source: WARHURST et al. (2009, tab. 4.15, p. 100).

Table A2. Average term-time income and expenditures ofScottish undergraduates (£)

£Percentage of

incomePercentage ofexpenditure

Average total income 5157 100 83Student loan 1430 28 23Informal housing

contribution163 3 3

Informal livingcontribution

290 6 5

Term-time earnings 1945 38 31Education-related

grants and bursaries759 15 12

Other 570 11 9

Average total expenditure 6230 121 100Housing costs 1116 22 18Living costs 3954 77 63Participation costs 957 19 15Child specific costs 203 4 3Other costs 110 2 2Income shortfall 1073 21 17

Source: WARHURST et al. (2009, tabs 2.4 and 3.4, pp. 24, 56).

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 413

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

seen as rather conservative. Available evidence (Table A2)suggests that the average income shortfall of Scottishunder-graduates is significantly larger each year, amounting to£1073. Unfortunately Warhurst et al. do not elaborateon how the income shortfall might be explained, buthere it is expected to be met by some combination ofunderreported informal contributions (within householdtransfers), earnings outwith term time (drawing onsavings) and commercial credit.

Warhurst et al. estimated the average term-timeemployment income of Scottish undergraduates at£1945. Here it is assumed that this average holds forincoming students from other parts of the UK, whileforeign students are assumed not to participate in thelabour market. Finally, the direct import content of stu-dents’ expenditure is deducted, which is assumed toequal that of Scottish households in general (32%), asreported in the Scottish IO tables (Table A3).

Having estimated the students’ net contribution tofinal demand, it is possible to estimate the ‘knock on’impacts of their consumption spending. A studentexpenditure vector estimated by KELLY et al. (2004)was used to derive the spending impact of the differentstudent groups in Scotland. In total they supportapproximately 0.40% of output (Table A4).

NOTES

1. For details of the construction of the IO accounts, thederivation of the income and expenditure structure of

the HEIs sector, and the data sources used, seeHERMANNSSON et al. (2010a).

2. For an overview, seeGOLDSTEIN (2009),HERMANNSSON

and SWALES (2010), and MCMAHON (2009).3. Indeed, the analysis may have a more general applica-

bility, since even where budgets are not devolved theremay be interest in identifying the demand-side impli-cations of expenditure switching.

4. For amore detailed account of themethodology of impactstudies and regional multipliers, see ARMSTRONG andTAYLOR (2000), LOVERIDGE (2004), and MILLER andBLAIR (2009).

5. The distinction between endogenous and exogenousactivity depends on the model and the application.Extended and dynamic IO models endogenize additionalelements of final demand, such as government expendi-ture and investment (BATEY and MADDEN, 1983;BATEY and ROSE, 1990; MADDEN, 1993). Also in theenvironmental accounting literature, MCGREGOR et al.(2004) suggested endogenizing trade in a single-regioncontext in an attempt to link local polluting emissionsfully to local private and public consumption.

6. The nature of the regional economy naturally governsthe acceptability of such assumptions. For example, inthe island economy of Jersey the institutional frameworkrestricts migration so that supply-side crowding out canbe expected even in the long run (LEARMONTH et al.,2007).

7. MCGREGOR et al. (2006) argued that although less fre-quently applied, the IO analysis is methodologicallysuperior to Keynesian income–expenditure models.However, the latter might be used in circumstanceswhere indicative results are considered sufficient or IO

Table A4. Impact of student spending in Scotland

Student origin

Scotland Rest of the UK Rest of the world Total

Output impact of student spending (£, millions) 353 141 215 708… as a percentage of gross output 0.20 0.08 0.12 0.40GDP impact of student spending (£, millions) 195 78 119 391… as a percentage of GDP 0.22 0.09 0.13 0.44Employment impact of student spending full-time equivalents 1721 687 1048 3456… as a percentage of Scotland employment 0.09 0.03 0.05 0.17

Note: GDP, gross domestic product.

Table A3. Derivation of per student spending

Location of domicile Scotland Rest of the UK Rest of the world

Gross average student spending (£) + 6230 7187 7187Income from employment (£) – 1945 1945Within household transfers (£) – 453Other income (£) – 570Income shortfall (£) – 1073Spending attributable to new commercial credit (£) + 346Exogenous average per student spending (£) = 2535 5242 7187Direct imports (£) (32%) – 816 1688 2315Net change in final demand per student (£) = 1719 3554 4872Number of students’ full-time equivalents × 114262 22052 24555Estimated net contribution to final demand by student population (£, millions) = 196.4 78.4 119.6

414 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

accounts are not available and cannot be constructed withthe resources available.

8. Except perhaps in the most recent studies based on theScottish IO Output tables. Although it should be noted thatthe multipliers for HEIs cannot be obtained directly fromthe Scottish IO tables but have to be disaggregated from theeducation sector based on supplementary information. Thefinal outcome of this process is sensitive to the exact processand the data sources. Some aspects of the HEIs’ incomesand expenditures are particularly important in this regard,such as the quality of the available data on imports.

9. For further details on Keynesian multiplier models, seeARMSTRONG and TAYLOR (2000), ch. 1.

10. The authors participated in a three-year-long UK-wideinitiative on the economic and social impacts of HEIs,for which the funders – the Economic and SocialResearch Council (ESRC), the Scottish FundingCouncil (SFC), the Higher Education Funding Councilof England (HEFCE), the Higher Education FundingCouncil of Wales (HEFCW), and the Department forEmployment and Learning Northern Ireland (DELNI)– emphasized regular public engagement in prearrangedknowledge-exchange seminars across the UK. Fordetails of the initiative, see http://ewds.strath.ac.uk/Default.aspx?alias=ewds.strath.ac.uk/impact/.

11. The precise way the formula works is given in more detailby FERGUSON et al (2003, 2007), though its operation inpractice is thought to be rather more flexible, so that whatis known as formula bypass can occur (CHRISTIE andSWALES, 2010; HEALD, 1994). This would be additionalfunding that extends the binding budget constraint anddoes not trigger the expenditure shifting discussed here.However, such formula bypass is typically linked to thefunding implications of decisions taken by the centralUK government on devolved budgets rather thanfunding outwith the block grant for additional expendi-ture decisions made by the devolved governments.

12. The Scottish Government does have limited powers tovary its expenditure through adjusting the standardincome tax rate up or down by 3 pence in the pound.This is the Scottish variable rate, but it has never beenused. For details, see, for example, LECCA et al. (2010)and MCGREGOR and SWALES (2005).

13. The public sector is aggregated from five sectors in theHEI-disaggregated IO table (IO115, IO116, IO117,IO118 and IO119). Approximately 10% of the sector’sfinal demand is from other sources than government.

14. For further discussion about analysing the impact ofexpenditure switching within an IO context, seeALLAN et al. (2007).

15. It should be noted that RUK exports include centralgovernment expenditures such as grants offered byresearch councils based on UK-wide competitivebidding. However, this category also includes incomefrom non-government sources, such as tuition feespaid by students from the RUK. Furthermore,income from intermediate demand is treated asendogenous in the IO model and is thereforecounted as part of the ‘knock-on’ impacts, ratherthan driving a ‘knock-on’ impact. For a discussion ofthis point and how it is sometimes misunderstood inimpact studies in practice, see OOSTERHAVEN andSTELDER (2002). For details of how these incomesare determined, see HERMANNSSON et al. (2010a).

16. For details, see Appendix A.17. WARHURST et al. (2009) built on and expanded the work

of CALLANDER et al. (2005).18. For the academic year 2008/2009, LÍN estimated the cost

of subsistence for obtaining a single European CreditTransfer System (ECTS) credit in Scotland was £142,where a full academic year would consist of sixty credits,amounting to anticipated costs of £8520 (see http://www.lin.is/Namslan/utlan/framfaerslutafla.html).

19. The category also includes support from private charities.Here the conservative stance is adopted that the charitiesare funded from Scottish contributions and therefore rep-resent a redistribution within the Scottish economy ratherthan an additional injection.

20. In principle, parents could be funding these transfers bydrawing on savings or taking out new credit, but it isassumed they are met with consumption switchingfrom parents to student.

21. Information on the composition of other income is notavailable in WARHURST (2009). Therefore, the conser-vative stance that it is non-additional to the Scottisheconomy is adopted here.

REFERENCES

ACS Z. J. (2009) Jaffe–Feldman–Varga: the search for knowledge spillovers, in VARGA A. (Ed.) Universities, Knowledge Transfer andRegional Development, pp. 36–56. Edward Elgar, Cheltenham.

ALLAN G., DUNLOP S. and SWALES K. (2007) The economic impact of regular season sporting competitions: the Glasgow Old Firmfootball spectators as sports tourists, Journal of Sport and Tourism 12(2), 63–97.

ANDERSSON R., QUIGLEY J. M. and WILHELMSSON M. (2009) Urbanization, productivity, and innovation: evidence from invest-ment in higher education, Journal of Urban Economics 66, 2–15.

ANSELIN L., VARGA A. and ACS Z. (1997) Local geographic spillovers between university research and high technology innovations,Journal of Urban Economics 42, 422–448.

ANSELIN L., VARGA A. and ACS Z. (2000) Geographic and sectoral characteristics of academic knowledge externalities, Papers inRegional Science 79(4), 435–443.

ARMSTRONG H. (1993) The local income and employment impact of Lancaster University, Urban Studies 30(10), 1653–1668.ARMSTRONG H. and TAYLOR J. (2000) Regional Economics and Policy, 3rd Edn. Blackwell, Oxford.BATEY P. W. J. and MADDEN M. (1983) The modelling of demographic–economic change within the context of regional decline:

analytical procedures and empirical results, Socio Economic Planning Sciences 17(5–6), 315–328.

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 415

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

BATEY P. W. J. and ROSE A. Z. (1990) Extended input–output models: progress and potential, International Regional Science Review13(27), 27–49.

BATTU H., FINCH J. H. and NEWLANDS D. (1998) Integrating Knowledge Effects into University Impact Studies: A Case Study of AberdeenUniversity. Department of Economics, University of Aberdeen, Aberdeen.

BLAKE C. and MCDOWELL S. (1967) A local input–output table, Scottish Journal of Political Economy 14, 227–242.BLEANEY M. F., BINKS M. R., GREENAWAY D., REED G. V. and WHYNES D. K. (1992) What does a university add to its local

economy?, Applied Economics 24, 305–311.BLUNDELL R., DEARDREN L. and SIANESI B. (2005) Measuring the returns to education, in MACHIN S. and VIGNOLES A. (Eds)What’s

the Good of Education?: The Economics of Education in the UK, pp. 117–145. Princeton University Press, Princeton, NJ.BRADLEY S. and TAYLOR J. (1996) Human capital formation and local economic performance, Regional Studies 30(1), 1–14.BROWN A. J. (1967) The ‘Green Paper’ on development areas, National Institute Economic Review May, 26–33.BROWNRIGG M. (1971) The regional income multiplier: an attempt to complete the model, Scottish Journal of Political Economy 18,

281–297.BROWNRIGG M. (1973) The economic impact of a new university, Scottish Journal of Political Economy 10(2), 123–139.BROWNRIGG M. and GREIG M. A. (1975) Differential multipliers for tourism, Scottish Journal of Political Economy 22, 262–275.CALLANDER C., WILKINSON D., MACKINNON K. and VEGERIS S. (2005)Higher and Further Education Students’ Income, Expenditure and

Debt in Scotland 2004–2005: A Report to the Scottish Executive. London South Bank University and Policy Studies Institute,London.

CHECCHI D. (2006) The Economics of Education: Human Capital, Family Background and Inequality. Cambridge University Press,Cambridge.

CHRISTIE A. and SWALES K. (2010) The Barnett allocation mechanism: formula plus influence?, Regional Studies 44(6), 761–775.FAGGIAN A. and MCCANN P. (2009) Human capital, graduate migration and innovation in British regions,Cambridge Journal of Econ-

omics 33, 317–333.FAGGIAN A., MCCANN P. and SHEPPARD S. (2010) Higher education, graduate migration and regional dynamism in Great Brita, in

VARGA A. (Ed.) Universities, Knowledge Transfer and Regional Development: Geography, Entrepreneurship and Policy, pp. 267–292.Edward Elgar, Cheltenham.

FERGUSON L., LEARMONTH D., MCGREGOR P. G., SWALES J. K. and TURNER K. (2003) The regional distribution of public expen-diture in the UK: an exposition and critique of the Barnett formula, in MØNNESLAND J. (Ed.) Regional Public Finance, EuropeanResearch in Regional Science 13, pp. 191–207. Pion, London.

FERGUSON L., LEARMONTH D., MCGREGOR P. G., SWALES J. K. and TURNER K. (2007) The impact of the Barnett formula on theScottish economy: endogenous population and variable population proportions, Environment and Planning A 39(12), 3008–3027.

FISCHER M. and VARGA A. (2003) Spatial knowledge spillovers and university research: evidence from Austria, Annals of RegionalScience 31(2), 303–322.

FLORAX R. (1992) The University: A Regional Booster? Avebury, Aldershot.GOLDSTEIN H. A. (2009) What we know and what we don’t know about the regional economic impacts of universities, in VARGA

A. (Ed.) Universities, Knowledge Transfer and Regional Development, pp. 11–35. Edward Elgar, Cheltenham.GREIG M. A. (1971) The regional income and employment effects of a pulp and paper mill, Scottish Journal of Political Economy 18(1),

31–48.HARMON C. and WALKER I. (2003) The returns to education: microeconomics, Journal of Economic Surveys 17(2), 115–153.HARRIS A., LLOYD M. G., MCGUIRE A. and NEWLANDS D. (1987) Incoming industry and structural change: oil and the Aberdeen

economy, Scottish Journal of Political Economy 34(1), 69–90.HARRIS R. (1997) The impact of the University of Portsmouth on the local economy, Urban Studies 34, 605–626.HEALD D. A. (1994) Territorial public expenditure in the United Kingdom, Public Administration 72, 147–175.HERMANNSSON K., LECCA P., LISENKOVA K., MCGREGOR P. and SWALES K. (2010c) The Importance of Graduates to the Scottish

Economy: A ‘Micro-to-Macro’ Approach. Strathclyde Discussion Papers in Economics Number 10-26. University of Strathclyde,Glasgow.

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. and SWALES K. (2010a) An HEI-Disaggregated Input–Output Table for Scotland.Strathclyde Discussion Papers in Economics Number 10-14. University of Strathclyde, Glasgow.

HERMANNSSON K., LISENKOVA K., MCGREGOR P. G. and SWALES K. (2010b) The Expenditure Impacts of Individual Higher EducationInstitutions (HEIs) and their Students on the Scottish Economy under Devolution: Homogeneity or Heterogeneity? Strathclyde DiscussionPapers in Economics Number 10-16. University of Strathclyde, Glasgow.

HERMANNSSON K. and SWALES K. (2010) Capturing the Overall Economic Impact of HEIs. Report commissioned by the TripartiteAdvisory Group (TAG) on Higher Education in Scotland (available at: http://www.scotland.gov.uk/Publications/2010/11/09155629/0).

HIGHER EDUCATION STATISTICS AGENCY (HESA) (2007) Resources of Higher Education Institutions 2005/06. HESA, Cheltenham.HM TREASURY (2003) The Green Book: Appraisal and Evaluation in Central Government. The Stationery Office (TSO) for HM Treas-

ury, London.JAFFE A. B. (1989) Real effects of academic research, American Economic Review 79(5), 957–970.KELLY U., MCNICOLL I. and MCLELLAN D. (2004) The Impact of the University of Strathclyde on the Economy of Scotland and the City of

Glasgow. University of Strathclyde, Glasgow.

416 Kristinn Hermannsson et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4

LEARMONTH D., MCGREGOR P., SWALES K., TURNER K. and YIN Y. P. (2007) The importance of the regional/local dimension ofsustainable development: an illustrative computable general equilibrium analysis of the Jersey economy, Economic Modelling 24,15–41.

LECCA P., MCGREGOR P. and SWALES K. (2010) Balanced Budget Government Spending in a Small Open Regional Economy. StrathclydePapers in Economics Number 10-20. University of Strathclyde, Glasgow.

LOVE J. and MCNICOLL I. (1988) The regional economic impact of overseas students in the UK: a case study of three Scottish uni-versities, Regional Studies 22(1), 11–18.

LOVE J. and MCNICOLL I. (1990) The economic impact of university funding cuts, Higher Education 19(4), 481–495.LOVERIDGE S. (2004) A typology and assessment of multi-sector regional economic impact models, Regional Studies 38(3), 305–317.LUNDVALL B. A. (2008) Higher education, innovation and economic development, in LIN J. Y. and PLESCOVIC B. (Eds) Higher Edu-

cation and Development, pp. 201–228. The World Bank, Washington, DC.MADDEN M. (1993) Welfare payments and migration in a nonlinear, extended input–output model with an application to Scot-

land, Papers in Regional Science 72(2), 177–199.MCGREGOR P. and SWALES K. (2005) Economics of devolution/decentralization in the UK: some questions and answers, Regional

Studies 39(4), 477–494.MCGREGOR P., SWALES K. andMCLELLAND. (2006) The overall impact of higher education institutions on regions: a critical review,

in Final Report of the HERE Network (available at: http://ewds.strath.ac.uk/impact/Background/StageOneNetworks.aspx).MCGREGOR P., SWALES K. and TURNER K. (2004) An Input–Output-Based Alternative to ‘Ecological Footprints’ for Tracking Pollution

Generation in a Small Open Economy. Strathclyde Discussion Papers in Economics Number 04-04. University of Strathclyde,Glasgow.

MCGREGOR P., SWALES K. and YIN Y. P. (1996) A long-run interpretation of regional input–output analysis, Journal of RegionalScience 36, 479–501.

MCGUIRE A. (1983) The regional income and employment effects of nuclear power stations, Scottish Journal of Political Economy 30,264–274.

MCMAHON W. (2009) Higher Learning, Greater Good. John Hopkins University Press, Baltimore, MD.MCMAHON W.W. (2004) The social and external benefits of education, in JOHNES G. and JOHNES J. (Eds) International Handbook on

the Economics of Education, pp. 211–259. Edward Elgar, Cheltenham.MCNICOLL I. (1981) Estimating regional industry multipliers, Town Planning Review 52, 80–88.MCNICOLL I. (1993) The Impact of Strathclyde University on the Economy of Scotland. University of Strathclyde, Glasgow.MCNICOLL I. (1995) The Impact of the Scottish Higher Education Sector on the Economy of Scotland. Committee of Higher Education

Principals (COSHEP), Glasgow.MCNICOLL I., KELLY U. and MCLELLAN D. (1999) Economic Aspects of Scottish Higher Education Institutions: Report to the Committee of

Scottish Higher Education Principals. Committee of Higher Education Principals (COSHEP), Glasgow.MCNICOLL I., KELLY U. andMCLELLAN D. (2003) The Economic Impact of Scottish Higher Education. Universities Scotland, Edinburgh.MILLER R. E. and BLAIR P. D. (2009) Input–Output Analysis: Foundations and Extensions, 2nd Edn. Cambridge University Press,

Cambridge.OOSTERHAVEN J. and STELDER D. (2002) Net multipliers avoid exaggerating impacts: with a bi-regional illustration for the Dutch

transportation sector, Journal of Regional Science 42(3), 533–543.OSBORNE M. J. (2004) An Introduction to Game Theory. Oxford University Press, Oxford.PARKER D. D. and ZILBERMAN D. (1993) University technology transfers: impacts on local and U.S. economies, Contemporary Econ-

omic Policy 11(2), 87–99.PSACHAROPOULOS G. and PATRIONS H. A. (2004) Human capital and rates of return, in JOHNES G. and JOHNES J. (Eds) International

Handbook on the Economics of Education, pp. 1–57. Edward Elgar, Cheltenham.SIEGFRIED J. J., SANDERSON A. R. and MCHENRY P. (2007) The economic impact of colleges and universities, Economics of Education

Review 26, 546–558.SWENSON D. (2006) Input–outrageous: the economic impacts of modern biofuels production. Paper presented at the Mid-

continent Regional Science Association and the Biennial Implan National Users Conference, Indianapolis, IN, USA, 8–10June 2006 (available at: http://www.econ.iastate.edu/sites/default/files/publications/papers/paper_12644.pdf).

VARGA A. and SCHALK A. (2004) Knowledge spillovers, agglomeration and macroeconomic growth: an empirical approach,Regional Studies 38(8), 977–989.

WARHURST C., COMMANDER J., NICKSON D., SYMEONIDES A., FURLONG A., FINDLAY J., WILSON F. and HURRELL S. (2009) Higherand Further Education Students’ Income, Expenditure and Debt in Scotland 2007–08. Scottish Government Social Research, Edin-burgh (available at: http://www.scotland.gov.uk/Publications/2009/06/24115743/02).

‘Policy Scepticism’ and the Impact of Scottish HEIs on their Host Region 417

Dow

nloa

ded

by [

Uni

vers

ity o

f A

rizo

na]

at 2

2:37

04

July

201

4