• A Introduction

  A Government could print a good edition of Shakespeare’s works, but it could not get them written.¹xA. Marshall, ‘The Social Possibilities of Economic Chivalry’, The Economic Journal, Vol. 17, No. 7, 1907, p. 22.

  Imprisonment is the most expensive sanction in the Western world. For instance, it costs the state 167€ per day to imprison an offender in Finland. On the other hand, the costs of supervising an offender at home using the electronic monitoring devise are only 60€, and imposing on the offender community service is around 14€ per day.²xCriminal Sanctions Agency, ‘Criminal Sanctions Agency Statements and Annual Report for the Year 2012’, 2013 (in Finnish), available at <www.rikosseuraamus.fi/material/attachments/rise/julkaisut-monisteetjaraportit/6FcvDvctb/1_2013_TP_ja_toimintakertomus_2012_korj220313VALMIS.pdf>, accessed on 12 February 2014; Criminal Sanctions Agency, ‘Electronic Monitoring’, 2014, <www.rikosseuraamus.fi/en/index/sentences/monitoringsentence.html>, accessed on 4 April 2014. For prison costs in different American states, see Vera Institute for Justice, 2012, <www.vera.org/pubs/special/price-prisons-what-incarceration-costs-taxpayers>, accessed on 9 June 2014. For a comparative view of costs in selected European prisons and the average costs of prison in the United States, see Table 1.³xPrisons are not only costly but also considered to have a criminogenic effect. Prisoners are isolated from their supporting environment (family, friends), associate with other criminals and turn to unproductive individuals who do not develop their working skills. Those circumstances potentially increase recidivism and, in turn, crime rates, harming even further the society. See e.g. P. Bayer, R. Hjalmarsson & D. Rozen, ‘Building Criminal Capital behind Bars: Peer Effects in Juvenile Corrections’, The Quarterly Journal of Economics, 2009, pp. 105-147, available at <http://qje.oxfordjournals.org/content/124/1/105.short>, accessed on 19 June 2012; K.M. Chen & J.M. Shapiro, ‘Do Harsher Prison Conditions Reduce Recidivism? A Discontinuity-Based Approach’, American Law and Economics Review, 2007, pp. 1-29, available at <http://aler.oxfordjournals.org/content/9/1/1.short>, accessed on 19 June 2012; P. Gendreau, C. Goggin & F.T. Cullen, ‘The Effects of Prison Sentences on Recidivism’, Public Works and Government Services Canada, Ottawa, 1999, available at <www.prisonpolicy.org/scans/gendreau.pdf>, accessed on 19 June 2012; D. Glaser & M.A. Gordon, ‘Profitable Penalties for Lower Level Courts’, Judicature, Vol. 73, No. 5, 1990, pp. 248-252; T. Weigend, ‘Germany Reduces Use of Prison Sentences’, in M. Tonry & K. Hatlestad (Eds.), Sentencing Reforms in Overcrowded Times: A Comparative Perspective, Oxford University Press, New York and Oxford, 1997, pp. 177-178 (hereinafter: ‘Sentencing Reforms’); P.J.P. Tak, ‘Sentencing and Punishment in the Netherlands’, in Sentencing Reforms, pp. 194-200, at 196; A. Ashworth, ‘The Decline of English Sentencing and Other Stories’, in M. Tonry & R.S. Frase (Eds.), Sentencing and Sanctions in Western Countries, Oxford University Press, New York and Oxford, 2001, pp. 62-91, at 67 (hereinafter: ‘Sentencing and Sanctions’); T. Lappi-Seppälä, ‘Sentencing and Punishment in Finland: The Decline of the Repressive Ideal’, in Sentencing and Sanctions, pp. 92-150, at 92. Prison costs are undisputedly associated with the number of incarcerated prisoners and the length of their imprisonment. Therefore, increasing prison population amplifies the need to search for cost-reducing policies. One direct method of decreasing such costs is to introduce alternative sanctions to custody. However, incarceration is an inevitable result for some groups of offenders, i.e. judgement proof and dangerous criminals. Hence, there is a need for policies which may affect the prison costs without abolishing this institution.

  Table 1 Estimated Imprisonment Costs in Selected Countries, Different Years (Per Prisoner)⁴xThe Criminal Sanctions Agency, ‘Criminal Sanctions Agency Statements and Annual Report for the Year 2012’, 2013 (in Finnish), p. 4, available at <www.rikosseuraamus.fi/material/attachments/rise/julkaisut-monisteetjaraportit/6FcvDvctb/1_2013_TP_ja_toimintakertomus_2012_korj220313VALMIS.pdf>, accessed on 12 February 2014 (Finland – average); M. Nellis, K. Beyens & D. Kaminski (Eds.), Electronically Monitored Punishments: International and Critical Perspectives, Routledge, Abingdon, 2013, pp. 150-153, 179-182 (Netherlands and Belgium – low security prison); P. Lindström & E. Leijonram, ‘The Swedish Prison System’, in P.J.P. Tak & M. Jendly (Eds.), Prison Policy and Prisoners’ Rights, Proceedings of the Colloquium of the IPPF, Wolf Legal Publishers, Nijmegen, 2008, pp. 559-570, at 563 (hereinafter: “Prisons’ Rights”) (Sweden – average of low and high security prison); A. Barbarino & G. Mastrobuoni, ‘The Incapacitation Effect of Incarceration: Evidence from Several Italian Collective Pardons’, IZA Discussion Paper No. 6360, 2012, p. 33, available at <http://ssrn.com/abstract=2010975>, accessed on 28 January 2013 (Italy); Prison Reform Trust, Bromley Briefings Prison Factfile, Prison Reform Trust, London, 2013, p. 73 (England and Wales – average of 12 private prisons); J. Bungerfeldt, ‘Electronic Monitoring in Sweden: The Past, Present and the Future’, CEP Conference Stockholm, 2012, slide 12 (Sweden – low security); M. Mellett, ‘Irish Prison Policy, Prison Regime and Prisoners’ Rights’, in P.J.P. Tak & M. Jendly (Eds.), Prison Policy and Prisoners’ Rights, Proceedings of the Colloquium of the IPPF, Wolf Legal Publishers, Nijmegen, 2008, pp. 407-427, at 411 (Ireland – average); C. Henrichson & R. Delaney, The Price of Prisons What Incarceration Costs Taxpayers, Vera Institute for Justice, New York, 2012, p. 10, available at <www.vera.org/pubs/special/price-prisons-what-incarceration-costs-taxpayers>, accessed on 3 April 2014 (United States – average and range. The lowest costs are in Kentucky, and the highest are in New York State).

  Country	Costs Per Day	Costs Per Year
  Netherlands	120€	43,800€ (2012)
  England and Wales	123€	45,060€ (2012)
  Belgium	126€	45,990€ (2012)
  Italy	127€	46,452€ (2004)
  Sweden	160€	58,400€ (2008)
  Finland	167€	60,843€ (2012)
  Sweden	235€	85,775€ (2012)
  Ireland	267€	97,700€ (2008)
  The United States	64€ (Average); 29€-119€ (Range)	23,187€ (2010) AV; 10,588€-43,559€ RA

  Source: Own table based on multiple sources.

  Imprisonment costs may be reduced by allowing the private market being involved in its management. The idea of private prisons⁵x‘Private prisons’ and ‘contracted prisons’ are used interchangeably in this article. has been widely discussed in the US context.⁶xJ. Dilulio, Jr., ‘What’s Wrong with Private Prisons’, Public Interest, Vol. 92, 1988, p. 66; C.H. Logan, Private Prisons: Cons and Pros, Oxford University Press, New York, 1990; R. Hardin, ‘Private Prisons’, Crime and Justice, Vol. 28, 2001, p. 265; D.E. Pozen, ‘Managing a Correctional Marketplace: Prison Privatization in the United States and the United Kingdom’, Journal of Law & Politics, Vol. XIX, 2003, p. 253. Yet, the attention given to this topic by the law and economics scholars is limited,⁷xK.L. Avio, ‘On Private Prisons: An Economic Analysis of the Model Contract and Model Statute for Private Incarceration’, New England Journal on Criminal & Civil Confinement, Vol. 17, 1991, p. 265; K.L. Avio, ‘The Economics of Prisons’, European Journal of Law and Economics, Vol. 6, 1998, p. 143; O. Hart, A. Shleifer & R.W. Vishny, ‘The Proper Scope of Government: Theory and an Application to Prisons’, The Quarterly Journal of Economics, Vol. 112, 1997, p. 1127 particularly in the European context.
  The contribution of the present article to the existing literature is fourfold. (1) The article provides a law and economic analysis of the market failures justifying public prisons as opposed to the supply of punishment by the free market. (2) It presents a comprehensive analysis of the bureaucracy and political science arguments in order to identify the sources of inefficiencies of publicly owned prisons. (3) To the best knowledge of the author, this article is the first to apply the principal-agent model to explain the potential inefficiencies in privately contracted prisons and offers possible solutions. In addition, based on the principal-agent model, Avio’s model of remuneration is extended. (4) Finally, contrary to the existing literature, this article expands the analysis to the European context as well and attempts to provide an explanation for the scarcity of prison privatization in the continental Europe.
  This article follows strictly the law and economics approach. The analysis focusses on the cost-effectiveness criterion. Therefore, alternative theories and other criteria are beyond the scope of the article.
  The term ‘prison privatization’ should not mislead. It does not refer to the transfer of prisons entirely to the free market. Rather, it denotes the practice of contracting with private parties to build and manage services inside prisons. The responsibility and the punishment administration remain under the government’s⁸xThe notions ‘government’ and ‘state’ are used interchangeably in this article. authority. There are different models of prison contracting. Those models range from full privatization, i.e. construction, ownership and operation of correctional facilities (e.g. the United States, United Kingdom), to limited outsourcing of different prison services (e.g. France).
  The article is structured as follows. Section B provides a brief overview of the current practice of private prisons in the United States, England and Wales and France. Section C analyses prison privatization from the law and economics perspective. It explores the sources of inefficiencies in the public and the private prisons and offers some solutions. An attempt to understand the scarcity of private prisons in Europe is provided in Section D. Lastly, concluding remarks are presented in Section E.

• B Private Prisons: Countries’ Experience

  Prison privatization is a potential method to reduce the costs of incarceration. Operating prisons by contracting out the services to private parties is not a novelty. It was practiced centuries ago in different countries around the world. After a period of stagnation and dominancy of the public sector in the ownership and operation of correctional institutions, privatization re-emerged in some countries.⁹xR.F. Culp, ‘The Rise and Stall of Prison Privatization: An Integration of Policy Analysis Perspectives’, Criminal Justice Policy Review, Vol. 16, 2005, p. 412.
  Prior to analysing the practice of public versus private prisons, this article provides a brief overview of the current experience in three jurisdictions. The United States is discussed since it was the first country to re-introduce the practice of involving the private market in the prisons’ construction and operation. The experience of England and Wales is important since they use private prisons to the largest extent in the whole Europe. Finally, the case of France is presented since it applies the closest model to prison privatization in the continental Europe.¹⁰xTo the best knowledge of the author, besides France, only Denmark, Prison Privatisation Report International, 2004, available at <www.psiru.org/justice/PPRI60.htm>, and one Land in Germany (Hessen) adopted to some extent the semi-private model of prison, at <www.serco.com/media/internationalnews/sercotorunfirstprivateprisoningermany.asp>, accessed on 18 March 2013. This overview assists in identifying the problems and evaluating the benefits of contracted prisons.
  In the United States, although some half houses and immigration detention centres were already contracted out during the 1970s, only in the mid-1980s, the first adult prison was privatized.¹¹xD.C. McDonald, ‘Private Penal Institutions’, Crime and Justice, Vol. 16, 1992, p. 361; Ministry of Justice (UK), Contracted-Out Prisons, 2014, available at <www.justice.gov.uk/about/hmps/contracted-out>, accessed on 23 May 2014. In 2012, the United States housed around 128,300 inmates in privately operated prisons, which constituted around 7% of the American prison population.¹²xL.E. Glaze & E.J. Herberman, Correctional Populations in the United States, 2012, Bureau of Justice Statistics, Washington DC, 2013, p. 10, available at <www.bjs.gov/index.cfm?ty=pbdetail&iid=4843>, accessed on 19 March 2014.
  The prison services are contracted out after a procedure of bidding.¹³xThe private company may own and operate the correctional institution, own and rent it to the state, or only provide prison services to publicly owned prisons, see McDonald 1992. The private bid has to offer at least the same tasks performed by the public prisons, yet for a lower price.¹⁴xHardin 2001. In Florida, for instance, when the private contracting commenced in the 1990s, the private bidders were expected to demonstrate that they are able to provide services at least as good as the public sector but with a reduction of at least 7% in costs.¹⁵xL. Lanza-Kaduce, K.F. Parker & C.W. Thomas, ‘A Comparative Recidivism Analysis of Releases from Private and Public Prisons’, Crime & Delinquency, Vol. 45, 1999, p. 28.
  The payment scheme for the services provided by the private prisons is typically per-prisoner-per-day. Thus, the revenue of the private companies depends on the number of prisoners and the length of incarceration. Generally, the state assures a minimum quota of prisoners.¹⁶xE.C. Viano, ‘America’s Prison System’, in Prisoners’ Rights, pp. 139-184. The contracts usually state certain standards the private companies are expected to meet (with regard to safety and rehabilitation programmes, for example). In addition, the contracts often require the presence of a public official on-site to monitor the performance of the private provider. Yet, in practice, it is rarely exercised due to the costs of monitoring.¹⁷xHardin 2001. Nevertheless, the contract is publicly available, hence enabling different interest parties to litigate and challenge the contract and the performance of the private sector.¹⁸xPozen 2003.
  Many attempts were made to assess the quality of the private prison management and the savings in costs. However, due to methodological difficulties, the results are inconsistent and point to different directions.¹⁹xLogan 1990; Hardin 2001; S. Carbral & S. Saussier, ‘Organizing Prisons through Public-Private Partnership: A Cross-Country Investigation’, Brazilian Administration Review, Vol. 10, 2013, p. 100; B.W. Lundahl et al., ‘Prison Privatization: A Meta-Analysis of Cost and Quality of Confinement Indicators’, Research on Social Work Practice, Vol. 19, 2009, p. 383; T.C. Pratt & J. Maahs, ‘Are Private Prisons More Cost-Effective than Public Prisons? A Meta-Analysis of Evaluation Research Studies’, Crime & Delinquency, Vol. 45, 1999, p. 358. Similar inconsistencies may be found in the empirical investigation of recidivism rates of inmates released from privately operated prisons.²⁰xLanza-Kaduce et al. 1999; W.D. Bales et al., ‘Recidivism of Public and Private State Prison Inmates in Florida’, Criminology & Public Policy, Vol. 4, 2005, p. 57; A.L. Spivak & S.F. Sharp, ‘Inmate Recidivism as a Measure of Private Prison Performance’, Crime & Delinquency, Vol. 54, 2008, p. 482.
  In England and Wales, the discussion about prison privatization re-emerged at the end of the 1980s. Following a debate, the Criminal Justice Act 1991, which regulated the bidding of contracts with the private sector for running newly opened remand prisons, was introduced. Subsequently, the law was expanded and starting from 1994, all prisons in England and Wales can be designed, constructed, financed and operated by the private sector.²¹xN. Panchamia, Competition in Prisons, Institute for Government, London, 2012, p. 1. As of January 2014, 14 prisons are privately managed in England and Wales.²²xMinistry of Justice 2014. In 2013, around 13,027 inmates were held in private prisons in England and Wales constituting 16% of the country’s prison population.²³xPrison Reform Trust, ‘Bromley Briefings Prison Factfile’, 2013, p. 72.
  The characteristics of contracts for private prisons in the United Kingdom differ from the United States. The Home Office provides considerably prescriptive contracts, with clear and measurable outcomes.²⁴xSee e.g. Prison Reform Trust, ‘Private Punishment: Who Profits?’, 2005. It poses higher demands for vocational and rehabilitation programmes than those practiced in public prisons. In order to increase accountability, the personnel in private prisons has to be authorized by the Home Office. In addition, supervisors from the state are present on-site to assure appropriate performance. They are also the authority to deal with disciplinary measures imposed on inmates. In case of not meeting the targets, the private providers may be penalized. There are examples over the years where the Home Office withheld performance-linked fees for non-satisfactory performance.²⁵xIbid.; Pozen 2003.
  One of the major criticisms against the British system of private contracting is the lack of transparency. In the United Kingdom, the private contracts enjoy from the ‘commercial-in-confidentially’ practice, thus being away from the public eye and not subject to scrutiny. The rationale of the state behind this secrecy is that companies have the right to protect their price base and other relevant features of their activity (e.g. performance and standards) from their competitors. Some anecdotal evidence of the extreme approach taken by the state may be found in the first assessment study of private prisons. In this study, the Home Office withheld the contract details even from the researchers it appointed to compare the efficiency of the first private prison to a public prison.²⁶xHardin 2001. This practice inhibits the possibility of private parties and organizations to challenge the contracts and the private prisons’ compliance.²⁷xPozen 2003.
  Due to the confidentiality, it is hard to understand the payment scheme practiced in England and Wales; however, it seems that a fixed price is paid in each contract. The fee is for full performance, and sums may be deducted from this fee in case of incomplete compliance with the contract. However, in effect, it is similar to the US model of per-prisoner-per-day fee.²⁸xHardin 2001; Pozen 2003. In addition, England and Wales piloted with an alternative payment scheme, i.e. the PbR model. For an explanation, see Section 3.3.3.
  The performance and costs of private prisons in England and Wales were measured on different occasions. It seems that most studies support the cost-reduction advantage of private prisons. Furthermore, some studies assert that the quality of services provided under contracts is at least as good as the public sector, and in some fields even better. Nevertheless, it seems that the security is better in public prisons.²⁹xG. Sturgess, ‘The Sources of Benefit in Prison Contracting’, in V. Helyar-Cardwell (Ed.), Delivering Justice: The Role of the Public, Private and Voluntary Sectors in Prisons and Probation, Criminal Justice Alliance, London, 2012, pp. 31-40; W. Tanner, ‘The Case for Private Prisons’, Reform Ideas, No. 2, 2013; Panchamia 2012.
  In continental Europe, exercise of force is regarded as a state monopoly (Rosenthal & Hoogenboom 1990).³⁰xU. Rosenthal & B. Hoogenboom, ‘Some Fundamental Questions on Privatisation and Commercialisation of Crime Control, with Special Reference to Developments in the Netherlands’, in Privatisation of Crime Control: Collected Studies in Criminological Research (Vol. XXVII), Council of Europe, Strasbourg, 1990, pp. 17-43. Therefore, prison privatization is not common, and delegation of significant custodial authority fails to exist. France adopted a private-public model that is the closest to prison privatization in the continental Europe.
  In 1987, France actually considered to adopt the American model of prison privatization by privately building and operating a new prison containing 13,000 beds (i.e. ‘The Programme 13,000’). However, a political discussion subdued this idea. Instead, a ‘hybrid management’ model was introduced. Under this model, the bidding involves only the building of the prison, yet the management remains the prerogative of the public sector. In addition, some ancillary services are contracted out, e.g. food, vocational programmes, health care, but the warden duties ought to be performed exclusively by public employees. The first hybrid prison was built in 1990, and by 2009 around 40 prisons operated under this model. In that year, 36% of the French prison population was accommodated in these prisons. The prediction for 2012 was that 50% of the prisons population would be accommodated in hybrid prisons or public prisons using private services.³¹xCarbral & Saussier 2013.
  The payment scheme in France is a fixed amount for the provided services, and above that, some expected profit (i.e. cost-plus contract³²xFor the explanation of the ‘cost-plus contract’, see P. Bajari & S. Tadelis, ‘Incentives versus Transaction Costs: A Theory of Procurement Contracts’, RAND Journal of Economics, Vol. 32, 2001, pp. 387-388.). If the prisons’ capacity increases, the state has to increase the paid amount. There are no incentives to reduce costs since all the expenses are covered by the state. For instance, the wages of the private sector workers are comparable to the public sector. Nevertheless, the quality of the services is higher since practices from private management are adopted.³³xCarbral & Saussier 2013. The lack of incentives and the higher quality led to higher costs of the hybrid prisons as compared to public prisons. In 2006, the French Court of Auditors compared the costs of hybrid prisons to comparable public prisons and found that in the years 1999-2003 the costs of hybrid prisons were between 8% and 33% higher than the public prisons.³⁴xCour des Comptes, Garde et Reinsertion: La Gestion des Prisons, Rapport Public Thématique, 2006, p. 175.

• C Private Prisons: The Law and Economics Perspective

  The first step into the analysis of contracted prisons³⁵xAs defined in the introduction. Namely, contracted-out prisons sponsored by the state, rather than a punishment entirely supplied by the private market. is to understand why we need publicly owned or sponsored prisons rather than supplying this good by the free market. Following this analysis, the present section provides a comprehensive application of the bureaucracy and political science arguments to explain the inefficiencies of publicly owned prisons and the benefits of privately contracted institutions. Finally, potential problems of prison privatization are discussed and dealt with using the law and economics instruments.

  I Public Prisons: Why State Intervention?

  In the economic literature, services and goods may be optimally allocated through the operation of a free market. The question arises then why prisons, a commodity which provides services, are public. The same literature states that intervention in the market is justified in the event of a market failure. The four main failures are as follows: (1) limitations to market competition, i.e. monopolies, (2) externalities, (3) public goods and (4) asymmetric information (Ogus 2004).³⁶xA. Ogus, Regulation: Legal Form and Economic Theory, Hart Publishing, Portland, 2004. At first glance, the strongest argument for state prisons is the public good notion.
  For a commodity or a service to be considered a public good, it has to have non-rivalrous consumption and non-excludability. The former means that consumption of a public good does not reduce its quantity for other consumers. The later, on the other hand, expresses that the provider of the good may not exclude (or it is too costly) non-paying consumers from using it. When a good is public all individuals may benefit equally from this commodity, regardless the question whether they paid for it or not. Therefore, there is an incentive to free ride on paying consumers. In such circumstances, there will be a shortage of suppliers.³⁷xR. Cooter & T. Ulen, Law and Economics (4th edn), Addison Wesley, Boston, 2003.
  Another way of looking at the public good problem is through positive externalities.³⁸xOgus 2004. The supplier of the good provides a service which may benefit people who did not pay for it. Thus, the private marginal benefit curve, facing the supplier, is lower than the marginal benefit curve of the society. As a result, the supply is lower than the social demand.
  The state may correct for this market failure in two ways. It may either provide the good by itself or alternatively, the state may subsidize private firms to supply the good. The most cited example for a public good is national defence. It provides protection for all citizens, whether they pay for the service or not.³⁹xCooter & Ulen 2003.
  Prisons may be regarded as a public good. The goal of this institution is to protect the society from crime by constraining the freedom of certain offenders and deterring other potential criminals from committing crimes.⁴⁰xOther goals might be to provide rehabilitation for offenders. Nevertheless, regardless of the methods used in different prisons to deal with prisoners, the end of all custodial sanctions is to protect the society from crime. This might be achieved through incapacitation, deterrence, rehabilitation, and so on. The benefits one group of citizens derives from the existence of prisons (protection from crime) do not reduce the quantity of those benefits for other individuals. In addition, non-paying consumers may not be excluded from enjoying the prisons’ benefits. Once the offenders are in prison, they may not harm the people outside irrespective of the question whether those people paid for the protection or not. This situation is expected to lead to free riding where not all consumers are paying for the good. Consequently, the private marginal demand facing the supplier would be lower than the social marginal demand and if provided by the private market, there would be a shortage of prisons. To correct for this market failure, the state may build and operate prisons by itself, using the resources collected from taxes. Indeed, most prisons in the Western countries are owned and managed by the state. However, governmental intervention may be minimized while still correcting for the market failure. The private market may provide the prisons and its services while being paid from the state budget to avoid shortage of prisons. This is the essence of prison privatization which is advocated in this article.

  II Possible Inefficiencies of State Owned and Operated Prisons

  Following the conclusion that prison is a public good and state intervention in this market is necessary, the next question arises whether public ownership and operation are efficient. There are several characteristics which may impede the efficiency of public owned prisons. Those features may be learned from the bureaucracy and political science literature.

  1 Politicians

  According to the public choice theory, politicians, as other rational individuals, maximize their utility function. Politicians’ main interest is to maximize their re-election odds.⁴¹xW.A. Niskanen, ‘Bureaucrats and Politicians’, Journal of Law and Economics, Vol. 18, 1975, p. 617; Ogus 2004; M. Boycko, A. Shleifer & R.W. Vishny, ‘A Theory of Privatisation’, The Economic Journal, Vol. 106, 1996, p. 309. Therefore, it is argued that politicians use the budget to promote their political goals rather than increase efficiency of the services they provide. Through transfers of benefits to their supporters, they assure their future position in the power. One case is the satisfaction of employees and labour unions. Politicians overinvest in employment in order to secure votes for future elections. This results in excessive employment and inefficient investment in remunerations. The explanation of such behaviour is that the benefits politicians derive from these actions are internalized (securing support). Yet, the costs of overinvesting in employment are mainly externalized to the Treasury and the public.⁴²xBoycko et al., Ibid.
  The easiest way for politicians to secure support is through publicly owned and operated corporations since they decide on the size of the employment and financial benefits which may be given through different projects.⁴³xA. Shleifer, ‘State versus Private Ownership’, The Journal of Economic Perspectives, Vol. 12, 1998, pp. 133-150. This problem may be illustrated in the following way: “politicians are likely to be more responsive to the interests of groups which benefit from productive inefficiency (employees, managers, and other input suppliers) than those which must bear the losses (taxpayers and consumers), because the financial stake of the former per individual is greater, and they are better organized”.⁴⁴xOgus 2004, p. 277. In other words, the collective action problem of taxpayers drives politicians to neglect the interests of this group and to promote on their expense other well-organized and influential interest groups. A good example for this behaviour may be found in the post-World War II Britain. The government at that time maintained the inefficient coalmine industry forasmuch as the miners’ union could bring down the government.⁴⁵xShleifer 1998.
  Another way to satisfy support is by hiring or nominating managers of the public institutions based on loyalty rather than merits.⁴⁶xS.O. Richardson, The Political Economy of Bureaucracy, Routledge, Abingdon, 2011; Logan 1990. Inevitably, this might preclude an efficient management since those managers may lack the skills and the proper incentives to operate the institutions efficiently. In addition, forasmuch as politicians are replaced every couple of years (or more in case of re-election), there might be a high turnover of the managers.⁴⁷xIf politicians are hiring based on loyalty, every politician would be interested in promoting his own supporters, thus replacing the old managers. This situation has a potential to result in loss of experience and instability in the institutional management.
  In the context of prisons, when the correctional institution is publicly owned and operated, politicians may over-invest in the employment of the prison staff, nominate managers who are loyal to them and provide excessive benefits (e.g. vacation, pensions). Those actions will potentially secure the future support and votes of the employees and the unions. Since labour costs constitute two-thirds of the prison expenditure,⁴⁸xHart et al. 1997; Sturgess 2012. the scope for inefficient resource allocation is significant and the ‘losers’ are the taxpayers. Indeed, in England and Wales, the source of savings in private prisons came mainly from the reduction of the staff members and their salaries/benefits. Since the private employees are not subject to the national payment rules, the providers usually adjust their wages to the market. In addition, employees in private facilities work longer hours and enjoy fewer benefits, e.g. holidays, pension, and so on.⁴⁹xTanner 2013.
  Other source of overinvestment might be entrenched in the conflicting goals and different costs politicians face. In his recent article, Gerrit de Geest presented a model of immunity where he demonstrated that officials who externalize precaution costs will over-invest in them if their actions are not subject to immunity.⁵⁰xG. De Geest, ‘Who Should Be Immune from Tort Liability?’, Journal of Legal Studies, Vol. 41, 2012, p. 291. The example provided by the author is that of fire fighters. A fire fighter needs to balance between the damage which may be caused due to excessive usage of water on the one hand and a fire damage of the property due to insufficient action on the other hand. In a situation where the fire fighter may be held liable for the water damage, avoiding this measure is a precaution. Since the precaution costs are mainly externalized to the society (a burning house), while the harm is internalized by the fire fighter, he is expected to choose to overinvest in precaution costs, i.e. to avoid action. This result explains why some officials have immunity from liability, e.g. police officers (p. 292). This situation may be translated to the context of prisons. Politicians need to balance between different goals, for instance, between imposing restrictive conditions on the prisoners in order to maintain safety on the one hand and maintaining those prisoners’ human rights on the other hand. Although they might be formally immune to liability for prisoners’ escape, in practice, they pay a very high political price for such events. There are examples from different countries where ministers had to resign from their public position following a publicized escape case of a convicted prisoner.⁵¹xFor example, following an escape of a notorious child-molester and a murderer Marc Dutroux, two Belgian ministers (the Minister of Interior and the Minister of Justice), along with the police chief, had resigned. See <www.bendevannijvel.com/andere/dutroux_inleiding.html>, accessed on 12 May 2014 (in Dutch). Another example may be found in Sweden where following the escape of a Swedish spy, the Justice Minister had resigned. See <www.nytimes.com/1987/10/20/world/a-swede-resigns-over-spy-s-escape.html>, accessed on 12 May 2014. See also Logan 1990, p. 71, asserting that public officials are more responsive to political loses rather than economic losses since the later can be externalized to the taxpayers. As a result, politicians will overspend on security. Moreover, some empirical evidence presents a ‘chilling effect’ resulting from similar events. For instance, one study investigated the effect of a negligence lawsuit against members of a state psychiatric hospital. The defendants were accused for negligently releasing a patient who later on committed a murder. The authors of the study found that following this lawsuit, the number of released patients had significantly decreased.⁵²xN.G. Poythress & S.L. Brodsky, ‘In the Wake of a Negligent Release Law Suit: An Investigation of Professional Consequences and Institutional Impact on a State Psychiatric Hospital,’ Law and Human Behavior, Vol. 16, 1992, p. 155.
  This situation might be analysed using the immunity model. The politician internalizes the costs of a prisoner’s escape. Yet, he may externalize the costs of precaution, i.e. the financial burden of over-restrictive measures on the prisoners and the non-tangible costs of violating their human rights. The financial costs are borne by the taxpayers. The costs of violating the inmates’ human rights are borne by the prisoners or by the taxpayers in case the court rules compensations in favour of the prisoner. Under these circumstances, the public officials are strongly incentivized to over-spend the prison budget on safety measures, even where it is excessive and inefficient.⁵³xFor example, in one maximum-security prison in the Netherlands, there are currently five prisoners and around 60 guards. This ratio seems excessive and inefficient. See H.G. van de Bunt et al., Gevangen in de EBI. Een empirisch onderzoek naar de Extra Beveiligde Inrichting (EBI) in Vught, Boom Lemma Uitgevers, Den Haag, 2013.

  2 Bureaucracies

  In 1967, Anthony Downs provided a definition of a bureaucrat, suggesting that one of his unique features is that his output may not be evaluated in the market. In the private market, consumers’ behaviour is a signalling device for the company to assess whether they act efficiently. If consumers are willing to pay a higher price for the firm’s output than the costs of its input, the firm knows its products are valued in the market. In this type of organization, employees may also be assessed based on their performance. In a bureaucratic organization, on the other hand, the products and services are not evaluated in the market. Thus, it is difficult to measure whether the financial burden the taxpayer carries in sponsoring the input of bureaucracy justifies the utility (output) he receives from the actions of this organization. Consequently, the budget or the income the bureaucratic organization receives is not related to the quality of its performance.⁵⁴xA. Downs, Inside Bureaucracy, Little, Brown and Company, Boston, 1967, pp. 25-30. So what then determines the budget?
  William Niskanen analysed bureaucratic behaviour through the public choice theory. He suggested that bureaucrats, like other individuals, maximize their utility. Their utility function includes the desire for salary, prerequisites, reputation and power. Those goals may be achieved through the increase of the bureau’s budget. Therefore, what bureaucrats maximize is the budget. As a result of this structure, Niskanen predicts that the output and the budget of bureau would exceed the output of a private firm that faces the same costs and demand.⁵⁵xW.A. Niskanen, Jr., Bureaucracy and Representative Government, Aldine Atherton, Chicago, 1971, pp. 38-39.
  Larger budgets increase the prominence of the agency and the influence of the bureaucrat. Consequently, bureaucratic organizations resist changes which may lead to the decrease of their budget and the importance of their offices. Innovation which may reduce the costs of the provided service is therefore unwelcomed, unless the office may retain the saved resources.⁵⁶xRichardson 1990, pp. 25-26; Downs 1967, p. 196. In the context of public prisons, this preference suggests that there are no incentives in seeking methods to reduce the prison costs. On the contrary, reduction of incarceration costs would decrease future budget for prisons; thus, it might be perceived as unbeneficial by the relevant public officials.
  The overspending might be controlled by limiting the budget and defining targets. However, due to different social goals and limited information in the hands of the ministries, it is hard to set clear targets for the bureaucratic agencies. In addition, increasing efficiency is not a necessary result of a limited budget. For instance, costs might be saved by reducing the quality of the good.⁵⁷xOgus 2004, pp. 277-278.
  Another source of inefficiency in the public sector might be explained by the structure of bureaucracies. One important method of improving the quality of services or goods in private companies is through innovation.⁵⁸xIbid., p. 286. The differences between the public and the private sectors might shed some light on the reasons why private providers are more prone to innovate than public.
  An economic goal of an organization is productive efficiency, i.e. minimizing production costs (maximizing output as compared to the input). The structure of the private market induces incentives to achieve this goal. First, shareholders are the residual owners of the firm. They have interest in increasing the value of the firm; therefore, they would impose pressure on the managers to maintain efficient production. In publicly owned entities, there are usually no shareholders who would pressure the officials to act efficiently. The taxpayers might be regarded as the residual owners of the public property. However, due to the collective action problem, i.e. too dispersed and bear too minor individual costs, they are not incentivized to intervene in the public policy. A second instrument to increase efficiency is the market control. Inefficient firms face the risk of takeovers which usually results in the replacement of the incompetent management. This kind of market control does not exist in the publicly owned and operated enterprises since they usually serve as a monopoly. Third, the remuneration of managers may be attached to the performance of the firm, thus increasing his incentives for achieving efficient production. Managers of public organizations do not have a financial interest in the performance of the organization. Their salaries are usually set by the Ministry and the Treasury. Fourth, inefficient corporations face the risk of going bankrupt. This may constitute a strong motivation to constantly improve. On the contrary, due to soft budget constraint public entities do not face the risk of being shut down due to bankruptcy. The managers of such institutions know they may be bailed out by the government. Thus, they are not strict enough with their fiscal discipline and effectiveness of spending.⁵⁹xJ. Kornai, E. Maskin & G. Roland, ‘Understanding the Soft Budget Constraint’, Journal of Economic Literature, Vol. 41, 2003, p. 1095; Logan 1990. Finally, market competition drives the private firms to innovate and improve their production efficiencies. Inefficient companies are forced to increase prices in order to cover their costs. As a result, consumers may change their preferences to substitute products and purchase them from other firms.⁶⁰xOgus 2004, pp. 272-273. In contrast, since the public entity is usually a monopoly, there is no competition and the ‘purchaser’ of the good may not turn to a different supplier in case of productive inefficiency.⁶¹xG. Tullock, The Selected Works of Gordon Tullock: Bureaucracy, Vol. 6, Library Fund, Inc., Indianapolis, 2005, p. 283.
  The performance of public prisons as compared to private prisons is a particular case of the above-mentioned analysis. In a situation of contracted-out correctional institutions, the private provider is the residual owner of the prison and this institution’s saved costs. Therefore, the managers would be pressured by the shareholders to increase productive efficiency. On the other hand, the residual owners of public prisons are the taxpayers. Since the ‘ownership’ is dispersed, and the individual cost of inefficient performance of a public prison is not high, taxpayers do not have strong incentives to act against the prison policy. Furthermore, soft budget constraint removes the risk of bankruptcy in public prisons, therefore, decreasing their incentives to operate efficiently. Finally, in the absence of prison privatization, public prisons are not subject to competition and are not incentivized to improve their performance. This argument relies on the yardstick competition notion.⁶²xA. Shleifer, ‘Theory of Yardstick Competition’, The RAND Journal of Economics, Vol. 16, 1985, p. 319. The monopolized public prison market has no benchmark of efficiency. As a result, the taxpayers may not evaluate whether this public organization operates efficiently or wasting their contributions. Moreover, these public services are irreplaceable; hence, there is usually no threat of closure of inefficient prisons.
  Prison privatization in England and Wales serves as a good example for the importance and the significant benefits of yardstick competition. One important advantage of private contracting which was observed in the United Kingdom is the ‘spill overs’ of competition. Besides having a stimulating effect on the private sector to innovate and improve performance through introduction of new technologies, competition had a positive externality on the public sector. It was suggested that following the bidding process in England and Wales, some of the public prisons improved their performance and reduced their costs. This process was further exacerbated by the Market Testing idea. In England and Wales, there are biddings which are opened both for the private and the public sectors. Through this process, some of the private prisons returned to the Prison Service responsibility. This proved that with clear targets and free competition, the public sector might also improve,⁶³xHardin 2001, pp. 333-334; Tanner 2013, pp. 12-13; Panchamia 2012, p. 5; Sturgess 2012, pp. 33-40. therefore, adding a new advantage of opening the prison market to private companies.
  An additional obstacle for innovation in the public sector is the hierarchical structure of the bureaucratic organization. If a low-ranked agent desires to offer an innovative change which might improve the performance of the organization, he needs to exert a significant effort in filling different reports and waiting lengthy periods for a decision from superiors. The larger is the organization, the higher is the number of the decision makers who need to approve this change. The multiplication of decision makers reduces the probability the change will be approved. Furthermore, bureaucratic organizations have a tendency to retain the status quo and express risk-averse behaviour since changes usually do not guaranty higher benefits than costs.⁶⁴xDowns 1967, pp. 160, 195. As a result of this structure, the low-ranked agents, who do not internalize the benefits of their innovation but bear the costs of offering it, are not incentivized to propose innovations.⁶⁵xHart et al. 1997, p. 1129.
  Furthermore, since there is no market evaluation of the employees’ performance, there is no clear indicator for the ‘right’ behaviour. Behaviour in bureaucratic organizations is not shaped based on outcomes. Nevertheless, there is a need to assure the agents are promoting the tasks of the bureau. Consequently, the focus is on the actions rather than on the outcomes. This explains why bureaucracies usually have a large number of rules and procedures the employees must follow in order to complete their tasks (Downs 1967).⁶⁶xDowns 1967, p. 59. Prescribed rules rather than clear targets with a space for discretion hinder the possibility to innovate since innovation means changing the behaviour in order to reach the same outcome using more efficient method.
  Although privatization of prisons may correct for some inefficiencies induced by the characteristics of a public organization, it faces its own problems. A paradox of prison privatization is nicely presented by Kenneth Avio: “private prisons are socially superior to public prisons because the former seek efficiencies in the drive to maximize profits; public prisons are superior to private prisons because they are not driven by the profit motive”.⁶⁷xAvio 1998, p. 150 (emphasis added). Therefore, the next section discusses the inefficiencies derived from the profit-maximizing goals of private prisons and presents possible solutions.

  III Potential Inefficiencies in Private Prisons and Suggested Solutions

  1 The Problem

  Prison privatization may be analysed through the principal-agent model.⁶⁸xFor the application of other economic models to prison privatization, see e.g. Avio 1998, using the ‘product quality model’; Hart et al. 1997, applying the ‘incomplete contract model’. This model was developed by Jensen and Meckling in the context of the theory of the firm and the relationship between managers and other stakeholders. According to their definition, when one or more persons (principals) contract with another person (agent) for the delivery of certain services, the former delegates some decision power to the latter. Forasmuch as both parties are utility maximizers and there is asymmetric information, it is expected that the agent will not always act in the best interest of the principal. There might be a ‘direct’ conflict of interests between the principal and the agent when the manager extracts money from the firm and reduces its value for other stakeholders. On the other hand, the manger might simply make non-pecuniary decisions which benefit him but not the firm. In order to minimize the departure of the agent from the principal’s interest, the latter may create a proper incentive scheme to align the agent’s interest with his own. Alternatively, he may incur monitoring costs to assure that agent’s decisions maximize his utility.⁶⁹xM.C. Jensen & W.H. Meckling, ‘Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure’, Journal of Financial Economics, Vol. 3, 1976, pp. 305, 308, 313.
  The principal-agent model may be applied in the context of prison privatization. The principal is the government, or the public as represented by the government, and the agent is the private provider. When contracting with a private company an operation of a prison, the state partially delegates its powers to the winner of the bid. The operator of correctional institutions maintains discretion regarding different decisions related to the inmates’ life in the facilities. Forasmuch as the principal and the agent in this context have different aims and both are expected to maximize their utility, a conflict of interests emerges. According to the deterrence theory, the state’s goal of using prisons is to deter potential and actual offenders from committing crimes and thus reducing the crime rate.⁷⁰xAs long as the marginal costs of crime prevention does not exceed the marginal benefits derived from it. See G.S. Becker, ‘Crime and Punishment: An Economic Approach’, Journal of Political Economy, Vol. 76, 1968, pp. 169, 207-209. Other goals of correctional institutions are incapacitation and rehabilitation of criminals. The latter is achieved when released prisoners do not re-offend. On the contrary, the aim of private providers of prisons, who are profit maximizers, is to ‘keep the business running’. In other words, private operators of custody have a financial interest in increased number of prisoners, whether they are new or returning (recidivists). This motivation is especially strong with the payment scheme of per-day-per-prisoner, as practiced in the United States, which attaches the scope of revenue to the quantity of prisoners and incarceration days. Indeed, one of the concerns of the opponents to prison privatization is that corporations are lobbying for harsher sentencing in order to increase the portion of people sentenced to custody.⁷¹xFor the discussion on the lobby argument and evidence for its limited practice, see Hardin 2001, pp. 278-283. However, assuming private providers do not have an influence on the sentencing procedure and the number of ‘new’ prisoners, they still may enhance or not reduce recidivism. This is one of the divergence points between the principal (the government) and the agent (the private provider). The government is interested in providing the inmates with different rehabilitation and vocational programmes in order to reduce their criminality and improve their prospects. The private firm, on the other hand, is interested in minimizing the number of provided programmes. They might be motivated by two rationales. First, the less programmes they provide, the larger is their residual profit. Second, assuming those programmes have some negative effect on recidivism, fewer programmes will result in higher reoffending rates. Thus, the number of admitted prisoners will increase and, in turn, the private prison will profit. On the other hand, if the government specifies the number of required rehabilitation programmes, due to asymmetric information, the private provider may be simply incentivized to reduce the quality of the programmes, which is harder to observe.
  Following the principal-agent model, possible methods to align the interests of the state and the private provider are either by introducing a monitoring system or creating an incentive scheme. In other words, the principal may either regulate the agent’s behaviour by setting detailed rules of action and supervising for compliance, or to assess the outcomes of the agent’s performance. The target outcomes may relate, for example, to recidivism rates, which signal to some extent the quality of the rehabilitation programmes. In addition, the acceptable quantity of riots and inside-prison violence may be limited since they usually serve as a good proxy for prison mismanagement.⁷²xHardin 2001, p. 289.

  2 Regulating and Monitoring Behaviour

  Monitoring in private prisons is already applied in practice, intensively in the United Kingdom (see Section II). Under such system, the contract first needs to prescribe measurable and observable targets and specify which programmes are expected to be provided by the private operator. Although the UK designs more prescriptive contracts, in the US contracts also may specify what exactly is expected from the providers. For instance, in Florida, the contract needs to stipulate that work and educational programmes ought to be provided.⁷³xLanza-Kaduce et al. 1999, p. 30. In the next step, on-site inspectors from the public sector are employed to assure the contract is performed satisfactorily, thus monitoring the behaviour. This system is not optimal for solving the principal-agent problem. First, it entails non-negligible pecuniary costs. It requires the employment of public personnel in every private prison. In the United States, those costs are the reason why inspectors on-site are scarce in some states even when prescribed by the contract.⁷⁴xHardin 2001, p. 307. Second, it might have negative non-pecuniary consequences. In order to monitor, the contract has to be clear about the services which ought to be provided. The reason is that only observable and measurable actions may be monitored. However, such contracts might impede the very essence of prison privatization. One of the impetuses to introduce prison privatization is to stimulate research and innovation. Development of new technologies and methods has the potential to improve the quality of prison practices, while reducing or keeping the costs the same. In theory, transferring prison operation to the private sector has the potential to induce those corporations to design innovative programmes that would be more successful in reducing criminality.⁷⁵xSome evidence of successful privatization in different fields may be found in the World Bank, Bureaucrats in Business: The Economics and Politics of Government Ownership, The International Bank for Reconstruction, Washington DC, 1995. Specifying the number or the type of rehabilitation programmes that should be provided might inhibit the innovation process.⁷⁶xIt seems that the prescriptive character of the British contracts indeed minimized innovation. Panchamia 2012. In 2011, the Ministry of Justice relaxed the strictness of the contracts in order to leave more space for the providers to decide on the methods to deliver their services. It would simply bind the private sector to use the same methods which are used by the public sector and which do not guarantee the reduction of criminality.

  3 Focussing on the Outcomes: The Incentive Scheme

  A better method, in terms of fostering improvement, to align the agent’s interest to the principal’s is to introduce an appropriate incentive scheme. To this end, the government needs to specify the desirable outcomes and to create the proper incentives to achieve them. Nevertheless, the private providers should have discretion in choosing the instruments to meet those stated targets.
  In order to design an incentive scheme, insights from corporate governance may be useful. This literature deals with the mechanisms corporations use to align the interests of managers with those of shareholders. One of those mechanisms is compensation schemes. The most effective financial method to align the interests is the stock-based compensations. Under this scheme, the manager owns some shares of the company; thus, the increase of the firm’s value directly increases the manager’s wealth. This practice has the highest sensitivity of pay performance.⁷⁷xK.J. Murphy, ‘Executive Compensation’, in R. Layard, O. Ashenfelter & D.E. Card (Eds.), Handbook of Labor Economics (1st edn, Vol. 3, Chapter 38), Elsevier, 1999, pp. 2485, 2519, 2527, 2531-2532.
  Compensation schemes may be used also to incentivize private prisons to promote the public interest in reducing re-offending rates. One theoretical model was offered by Kenneth Avio. The author suggested linking the remuneration of private prisons to the recidivism level by including two elements in the payment scheme. The first part is a ‘flat-rate per-diem payment’ which would cover the daily operational costs. The second share of the compensations would resemble a royalty payment and would be paid in addition to the first part. Under the royalty payment, the private provider would receive a compensation for every period the offender does not recidivate. This payment may be revoked once the offender re-offends. The size of the royalty ought to be determined by a bid. Avio asserts that this scheme not only has the potential to improve the programmes provided in private prisons but might also induce the private providers to assist prisoners after their release.⁷⁸xAvio 1991, pp. 285-286.
  An example of compensation attached to results may be also found in practice. Recently, England and Wales piloted with a new payment system which is similar to the incentive-based scheme discussed in this article. The first pilot of this scheme, termed Payment by Results (PbRs), in private prisons was introduced in the HMP Doncaster⁷⁹xSee <www.justice.gov.uk/contacts/prison-finder/doncaster>, accessed on 23 March 2014. in 2011. According to this model, the private provider places at risk 10% of his annual revenue in order to re-assure a target reduction in the recidivism rate. In case the private prison does not meet the target, it losses the 10% of its revenue. Thus, the provider is ‘punished’ for not reducing re-offending rates. The assessment of the results is made through a binary process: if the released prisoner does not commit a crime in the following 12 months after his release for which he is convicted, the provider is not losing part of his revenue.⁸⁰xM. Chambers, ‘Expanding Payment-by-Results: Strategic Choices and Recommendations’, Policy Exchange, Vol. 1, 2012, pp. 11-12. The PbR was also tested in the public sector and with other services. Other piloted models were paying bonus on top of cost coverage for reducing recidivism; or paying no income unless the expected results are met (p. 11). In the Doncaster prison, the reduction target of recidivism rate was at least 5% as compared to the reconviction rate in the base year 2009.⁸¹xAn official response from the UK Ministry of Justice to the author’s inquiry (11 April 2014).
  Using ‘punishment’ to incentivize the reduction in re-offending rates as practiced in England and Wales might impose some difficulties. The corporate governance literature for instance points out that even though stock price is the right measure to assess the CEOs’ behaviour and decision, it might be ‘noisy’. Namely, the stock price might fall despite proper decisions made by the manager if the market experiences some financial crisis. To compensate the managers for this risk, the firm needs to pay a ‘risk premium’.⁸²xJ.E. Core, W.R. Guay & D.F. Larcker, ‘Executive Equity Compensation and Incentives: A Survey’, Economic Policy Review, Vol. 9, 2003, pp. 27, 30. Similarly, in the context of private prisons, the recidivism rates might be beyond the control of the private provider. One reason for such a result might be the lack of legal opportunities available to the released offender. Therefore, ‘punishing’ private providers rather than rewarding them for reducing recidivism rates might increase the risk transferred to those companies.
  Therefore, this article suggests using rewards, yet to further develop the compensation model offered by Avio. Differently from Avio, this article offers to set the ‘royalty’ payment not through a bidding process but as a percentage of the government’s saved costs from the reduced recidivism. This mechanism, similarly to the stock shares which attach the manager’s profits to the value of the company, will attach the profit of the private provider to the savings of the state. In other words, the reduction of recidivism would lead to a lower rate of crimes. Since crime imposes costs on the society and the state, less crime, means fewer expenses for the state. If the revenue of the private provider is attached to the government’s savings from reduced recidivism, the larger are the savings for the public, the greater is the profit for the private prison.
  Literature suggests that the majority of inmates inhabiting prisons are returning criminals.⁸³xA. Blumstein et al. (Eds.), Criminal Careers and “Career Criminals”, National Academy Press, Washington DC, 1986, p. vii; In 1994, around 70% of prisoners in the United States were rearrested within three years after their release. Around 52% returned to prison during this period, see T. Hughes & D.J. Wilson, Reentry Trends in the United States, Bureau of Justice Statistics, 2002, pp. 20-21. All the more so in Europe where first-time offenders are usually sentenced to alternative sanctions, and custody constitutes a last resort to manage recidivists and dangerous criminals. Therefore, recidivism imposes significant costs on society and rehabilitating prisoners have the potential to reduce crime. There are different studies measuring the saved costs to society from reducing crime.⁸⁴xFor instance, S.D. Levitt, ‘The Effect of Prison Population Size on Crime Rates: Evidence from Prison Overcrowding Litigation’, The Quarterly Journal of Economics, Vol. 111, 1996, p. 319; K.E. McCollister, M.T. French & H. Fang, ‘The Cost of Crime to Society: New Crime-Specific Estimates for Policy and Program Evaluation’, Drug and Alcohol Dependence, Vol. 108, 2010, p. 98. Those measures might assist the state to introduce a calculation of the avoided costs that occur due to the reduction of re-offending by released offenders. Paying the private providers, a portion from those savings has the potential to align their interest with the public desire to reduce recidivism.
  The suggested compensation scheme resembles the equity incentives provided to managers in corporations. In the context of corporate governance, the shareholders do not possess ex-ante the information on the decisions and choices which would maximize the firm’s value. Therefore, they delegate the decision-making power to managers who are believed to have better information. Yet, at the same time, the shareholders tie the manager’s wealth to their own, assuring he would make decisions that promote their interests.⁸⁵xCore et al. 2003, p. 32. In the prison privatization context, one of the rationales to adopt this approach is the belief that private profit-maximizing firms would use innovative methods and new technologies to manage prisons. The state may not ex-ante know which are the most cost-efficient methods. Possession of such information would obviate the need in contracting with private companies. Therefore, the state delegates the decision power to the private provider trusting he has better information and potential to improve prison operation. However, proper incentives need to be in place in order to assure the firms are providing the results which are expected from them. Paying the private prisons a portion from the saved costs due to reduced recidivism links the providers’ revenue to the main public interest, i.e. reducing crime and the associated costs. In turn, this compensation scheme has the potential to promote innovation in rehabilitation programmes and introduction of new methods to reduce the criminality of prisoners.
  The period for the reward should be limited in order to minimize the state’s expenses. Offenders usually re-offend soon after their release.⁸⁶xLanza-Kaduce et al. 1999, p. 34. If the private prison succeeded to reduce the criminal inclination of the offender, it would be depicted in the proximity of his release date. Thus, continuing paying for the prisoners’ choice not to re-offend in further periods as suggested by Avio would lead to wasted resources.⁸⁷xIt does not seem reasonable to believe that the private providers may invent a method which reduces criminal inclination only for a limited period, after which the offender would always return to crime. No doubt, in practice this might occur if, e.g. the legal opportunities of the released offender change after years. In this case, the benefits of crime may once again overweight its costs. However, the private prisons have no control over such situations.
  An additional potential contingency area which may induce the principal-agent problem is the choice of a method to reduce prison costs. There are some arguments that in the United States the costs were reduced at the expense of a lower quality of the prison management.⁸⁸xCarbral & Saussier 2013, p. 107. However, overall it does not appear that private contractors worsened the conditions in prisons, and some even improved it.⁸⁹xHardin 2001; Lundahl et al. 2009. In the United Kingdom, for instance, the main source of costs saving is through reduced prison personnel and their benefits. This element usually constitutes the largest portion of custodial expenditure (around two-thirds). Thus, employing less workers, younger on average and with more flexible recruitment and firing conditions significantly reduces the costs of personnel as compared to the public sector. In addition, in public prisons, it is harder to reduce the salaries and the different benefits the workers receive, hence giving the private prison advantages in this area.⁹⁰xTanner 2013, pp. 15-17. This practice is occasionally criticized. For instance, in 2003, the average annual wage of a prison officer in a public prison in the United Kingdom was £23,071 as compared to £16,077 in private prisons. The opponents of private prisons see this practice as inappropriate method to reduce costs. They assert that the less experienced private officers and the high turnover affect negatively the quality of the prison management.⁹¹xPrison Reform Trust 2005, p. 8. The validity of this argument is questionable since it is not proven that the number of employees and the benefits they receive in public prisons are optimal.⁹²xFor example, the high-security prison in the Netherlands currently accommodates five prisoners and around 60 members of the staff. The high number of the prison employees may not even be justified by the reasoning that they are required in order to prevent violence of the prisoners. The internal guidelines of this prison state that the prison officers are not allowed to face more than one prisoner at the same time. Thus, in case of a violent event of more than one prisoner, the staff is prohibited from intervening. The costs of such a prison seem exceeded and doubtful if the number of personnel may be considered optimal (van de Bunt 2013). In fact, the inefficiencies and the costs of the public prisons constitute some of the factors leading to prison privatization. In addition, increased employment and excessive benefits are often the result of inefficient political behaviour, as explained in Section B.
  Although reducing the prison staff may be a reasonable method to decrease costs, it seems that private prisons have a large potential for efficient costs reduction through innovation. The most pronounced innovation in private prisons is the installation of CCTV cameras, magnetic cards and privacy locks. In addition, some prisons in the United Kingdom introduced changes in the prisoner-staff relationship, i.e. using the prisoners’ name, which increased the sense of respect.⁹³xSturgess 2012, p. 34. However, further innovation may be enhanced by rightly designing contracts. For instance, bundling the prison construction with prison management might incentivize innovation in the structure of the prison.⁹⁴xA counter argument may be that separating the ownership from management might improve the accountability. In the United Kingdom, for instance, through the Market Testing procedure, the management of the prisons may be returned to the hands of the public sector if it outperforms the private provider. This is hard to do if the prison is owned by the private party and the buyback period is long, e.g. 20-40 years in the United Kingdom (Hardin 2001, p. 323). A possible way to resolve this issue is by requiring the private firms to agree in advance for a rent contract in case they underperform, and the management is transferred to the public sector. This provision in the contract may in fact add incentives to perform better than the public sector. Another solution is to provide a contract for construction and management but to maintain the ownership in the hands of the state. This way the private provider is still incentivized to produce an innovative design, yet there is no impediment for the Market Testing since the facility is owned by the state. Private providers may choose, for instance, the materials and design which would allow for savings in the electricity (e.g. cells which keep the warmth in the winter and are cool during the summer), or an innovative structure which would reduce the costs of monitoring. Jeremy Bentham’s Panopticon is a good example of structuring a prison with the vision of improving prison control while reducing costs.⁹⁵xM. Bozovic (Ed.), Jeremy Bentham, The Panopticon Writings, Verso, London, UK, 1995. The letters were written in the 18th century. The private companies building the prison for their own use may introduce similar innovations. A competition may enhance the incentives of private firms to modernize the prison structure. One possibility is to ex-ante require an innovative design in the stage of bidding. Another possibility is to reward financially, following a comparative assessment, the private prison with the most ground-breaking structure which led to cost reduction ex-post.
  In order to ensure the respect for human rights and basic conditions⁹⁶xNot overcrowded prisons, proper nutrition, good health care, and so on. in the private prisons, this article suggests combining the American and the British systems of accountability. On the one hand, the contracts and the private prisons’ performance should be transparent like in the United States. Individual prisoners and different interest groups should have access to courts in order to be able to challenge the prison management and secure prisoners’ rights. This type of a system opens the possibility to introduce reputation as an additional incentive mechanism to perform well through the media coverage.⁹⁷xIt might be expected that repeating negative coverage regarding violation of human rights in private prisons would lead to a public pressure to cancel the particular contract or to abolish prison privatization in general. In addition, as practiced in the United States, contracts should face a reliable threat of cancellation in case of gross violations. This way, the firms would be incentivized to maintain the required conditions in prisons in order not to lose the current and future contracts. On the other hand, to minimize even more the principal-agent problem, a stronger monitoring system should be adopted similar to the United Kingdom.⁹⁸xPozen 2003, pp. 277-280. However, in order to save costs and to avoid the reduction of impartiality of the controller due to close relationship with the management, the monitoring should be sporadic. In other words, an independent public official should make random and unannounced visits to the private facilities where he would inspect the conditions and conduct interviews with prisoners to assess the management. Each year, one private prison which outperforms all other prisons in respect of service quality and human rights protection should be financially rewarded to induce competition and incentivize increased quality. It should be noted that the need for monitoring does not increase the costs of private prisons as compared to public prisons, since the latter requires monitoring as well.⁹⁹xLogan 1990, p. 204.

• D Why the Scarcity of Private Prisons in Europe? Possible Explanations

  If prison privatization has the potential to overcome public organizations’ inefficiencies, the question arises, why is it not widespread in the continental Europe? It seems that two plausible reasons for resisting prison privatization in Europe are the constitutionality or the morality argument and the objection of interest groups.
  One of the main criticisms against privatizing prisons is that morally this is an inherent function of the state and may not be delegated to private parties. Opponents of private prisons assert that transfer of such power undermines the legitimacy of exercising punishment against offenders.¹⁰⁰xDilulio 1988, pp. 79, 81; S. McConville, ‘The Privatisation of Penal Services’, in Privatisation of Crime Control, Collected Studies in Criminological Research (Vol. XXVII), Council of Europe, Strasburg, 1990, pp. 77, 94. Although commentators both in Europe and in the United States pronounce this argument, the discussion over the morality of privatization has faded away in the United States yet remains strong in Europe.¹⁰¹xHardin 2001, p. 278. For instance, Section 124 to the Finnish Constitution states that “[…] a task involving significant exercise of public powers can only be delegated to public authorities”.¹⁰²xThe Constitution of Finland, 11 June 1999 (731/1999, amendments up to 1112/2011 included). This section is the reason why Finland does not privatize prisons.¹⁰³xT. Lappi-Seppälä, ‘Imprisonment and Penal Policy in Finland’, in P. Wahlgren (Ed.), Scandinavian Studies in Law, Vol. 54, Stockholm Institute for Scandinavian Law, Stockholm, 2009, pp. 333, 344; See also McConville 1990, p. 83, stating that even in jurisdictions where there is no explicit prohibition in the constitution for private administration of penal institutions, it might be implied. Another example might be found in France. This is the country which gotten the closest to privatize its prisons (‘semi-privées’) in the continental Europe, yet the exercise of custodial powers remains exclusively in public hands.
  Nevertheless, this objection might be challenged on several grounds. First, the definition of prison privatization should be clear and not misleading. As explained in the introduction, this process does not transfer the execution of the imprisonment punishment to the free private market. Rather, it allows the government to hire private parties to manage the correctional institutions on behalf of the state. In other words, the responsibility to set the goals, standards and legality of the prison remains under the public authority. The prisoners are state prisoners.¹⁰⁴xHardin 2001, p. 266. Under these circumstances, the government is accountable to the public for all the wrongs that might occur in the private prisons. The responsibility to choose a reliable private provider, to design the right contract and to supervise the results lies on the government and prison privatization does not change this status. Therefore, contracting with private firms to provide prison services should not be perceived as a transfer of significant power to non-public authority.
  The above-mentioned understanding of the definition sets the grounds for the second argument. The managers and workers who build and run the prisons are all hired rather than elected, whether they are public or private. Based on the economic theory, all individuals are rational and maximize their own utility. There is no reason to believe that civil servants are by nature more noble than the contracted workers and that they will promote the public interest. It is true that private providers are motivated by the profit interest. However, as discussed in Section B, agencies in public institutions are also motivated by self-interests. For instance, managers of agencies are often appointed based on their loyalty to the politician rather than based on their merits.¹⁰⁵xRichardson 1990, p. 27. Thus, not the public interest will steer the course of actions of this agency, but the interests of the politician. Consequently, it is not evident why prisons should be run only by state employees.
  Furthermore, the legitimacy of the prison and the limitation of power exercised in these institutions lie within the law. Therefore, the same rules guide the behaviour of all prison employees, whether they are civil or private. This argument is nicely illustrated by the following statement: “[f]or actors within either type of agency, it is the law, not the civil status of the actor, that determines whether any particular exercise of force is legitimate”.¹⁰⁶xLogan 1990, p. 54. Supporters of the constitutional argument are concerned that the profit incentive would lead to abuse of power or mismanagement of the prison. Yet, misconducts in prisons are not the prerogative of private providers. There are examples of abuse of power, mismanagement, suicides, violence and escapes in public prisons, as much as in private.¹⁰⁷xSee e.g. escapes and sexual abuse by public workers in preventive detention (TBS) in the Netherlands, available at <www.sevendays.nl/artikel/124986>, <www.misbruikdoorhulpverleners.nl/tbsgog&tbs.html>, accessed on 21 May 2014 (in Dutch); overcrowding and mal conditions in Italian prisons, available at <http://espresso.repubblica.it/inchieste/2014/04/25/news/carceri-ecco-perche-siamo-la-vergogna-d-europa-1.162997>, accessed on 21 May 2014 (in Italian). Moreover, private providers of prison services have strong incentives to avoid mismanagement and abuse of power since they perform in a competitive environment rather than acting as a monopoly (like state prisons). Excessive cost savings on the expense of quality and abuse of power is expected to lead to riots, lawsuits by inmates or human rights groups, and so on. Therefore, the private prisons risk losing future and even current contracts. In fact, the monopolistic power of state prisons might be one of the reasons why they are mismanaged.¹⁰⁸xLogan 1990, pp. 55, 75. Without the threat of being replaced, the incentives to improve are weak. As a result the profit-maximizing nature of contracted prisons increases the power of reputation as a safeguard for proper quality and care for the rights of prisoners.
  Another possible explanation for the resistance to private prisons in continental Europe is the objection of interest groups. One of the most relevant groups which have an interest to hinder this change is the labour union. Contracting with private providers to run the prisons means that public workers may expect to lose their employment and their benefits. Therefore, the labour union has a strong incentive to prevent privatization in this field. In fact, trade unions around the world are strong opponents of privatization in general since they do not want to lose the benefits they receive for their support (Shleifer 1998).¹⁰⁹xShleifer 1998, pp. 16-17. In the United States, the primary opponent of prison privatization was the American Federation of State, County and Municipal Employees (AFSCME). The reason was that this reform threatened their employment and power. Due to the union’s influence on state policy, private contractors focussed on states where the union was weaker.¹¹⁰xLogan 1990, pp. 11-12. Resistance of interest groups to private prisons exists also in Europe. The objection is raised by the European Public Service Union (EPSU) and by national unions as well.¹¹¹xSee e.g. the objection of the European Public Service Union (EPSU) to prison privatization in Denmark and in the Netherlands. One of their explicit arguments for the objection is that the staff would receive lower salaries and lesser benefits. Available at <www.mijnvakbond.nl/Waarschuwing-tegen-privatisering-gevangenissen?referrer=1133> (in Dutch) and at <www.epsu.org/a/7376>, accessed on 12 May 2014; similar resistance by the trade union to private prisons may be found in Belgium, available at <www.gva.be/archief/guid/acod-categoriek-tegen-privatisering-gevangeniswezen.aspx?artikel=acfadfd4-df43-4fd3-a136-9a8d58b8836a>, accessed on 12 May 2014 (in Dutch); in France, the National Union of Prison Directors (SNDP) also opposed contracting out prison services to private parties, available at <www.liberation.fr/societe/2012/01/31/prisons-gare-a-la-privatisation-du-service-public-penitentiaire_792402>, accessed on 12 May 2014 (in French). Therefore, the resistance of public service unions may be an additional plausible explanation why prison privatization in not widespread in the continental Europe. The only European country which significantly practices prison contracting is the United Kingdom. Interestingly, this reform was introduced during Margret Thatcher’s administration which substantially decreased the influence of labour unions on public policy.

• E Concluding Remarks

  The high costs of prisons and the need to reduce the enforcement system expenditure in Europe call for a legal reform. One possible cost-effective solution to incarceration may be prison privatization. Since custodial punishment is inevitable for certain offender population, pursuing ways to make this sanction less costly is important as much as searching for alternatives to replace it. It seems that the current use of private providers to construct and manage prisons is limited, and there is a room for reforming this practice. Countries should not be deterred by the private prisons solely based on the ideological argument that restriction of liberty is a matter of state. Outsourcing the prison management to private parties does not mean surrendering the power and the obligation of the state for those prisoners. The state remains involved and accountable for the offenders, and it is plausible to design carefully the private contracts in a way to assure human rights and proper conditions for the prisoners. It should be remembered that public prisons are also not immune to abuse of power and cases of violation of human rights. Therefore, the question should not be who in practice manages the correctional institutions, but what are the incentives and the degree of accountability in those prisons.
  The current experience with private prisons offers some insights to the cost-effectiveness of this practice, yet more robust research is needed to compare private with public providers. In general, there is a potential to reduce costs without tempering the quality of the prison services. However, there is a space for improvement, mainly in the context of incentives provided to the private prisons. Contracts should take into account and mitigate the principal-agent problem. Proper incentives ought to be introduced in order to align the interests of the public represented by the government and the private providers. Through such reforms of the current legal practices, the cost-effectiveness of private prisons may increase.

Noten

• * I would like to thank Michael Faure, Paul Mevis and Jaroslaw Kantorowicz for their valuable comments and support. All possible mistakes remain, however, my own.

• 1 A. Marshall, ‘The Social Possibilities of Economic Chivalry’, The Economic Journal, Vol. 17, No. 7, 1907, p. 22.

• 2 Criminal Sanctions Agency, ‘Criminal Sanctions Agency Statements and Annual Report for the Year 2012’, 2013 (in Finnish), available at <www.rikosseuraamus.fi/material/attachments/rise/julkaisut-monisteetjaraportit/6FcvDvctb/1_2013_TP_ja_toimintakertomus_2012_korj220313VALMIS.pdf>, accessed on 12 February 2014; Criminal Sanctions Agency, ‘Electronic Monitoring’, 2014, <www.rikosseuraamus.fi/en/index/sentences/monitoringsentence.html>, accessed on 4 April 2014. For prison costs in different American states, see Vera Institute for Justice, 2012, <www.vera.org/pubs/special/price-prisons-what-incarceration-costs-taxpayers>, accessed on 9 June 2014.

• 3 Prisons are not only costly but also considered to have a criminogenic effect. Prisoners are isolated from their supporting environment (family, friends), associate with other criminals and turn to unproductive individuals who do not develop their working skills. Those circumstances potentially increase recidivism and, in turn, crime rates, harming even further the society. See e.g. P. Bayer, R. Hjalmarsson & D. Rozen, ‘Building Criminal Capital behind Bars: Peer Effects in Juvenile Corrections’, The Quarterly Journal of Economics, 2009, pp. 105-147, available at <http://qje.oxfordjournals.org/content/124/1/105.short>, accessed on 19 June 2012; K.M. Chen & J.M. Shapiro, ‘Do Harsher Prison Conditions Reduce Recidivism? A Discontinuity-Based Approach’, American Law and Economics Review, 2007, pp. 1-29, available at <http://aler.oxfordjournals.org/content/9/1/1.short>, accessed on 19 June 2012; P. Gendreau, C. Goggin & F.T. Cullen, ‘The Effects of Prison Sentences on Recidivism’, Public Works and Government Services Canada, Ottawa, 1999, available at <www.prisonpolicy.org/scans/gendreau.pdf>, accessed on 19 June 2012; D. Glaser & M.A. Gordon, ‘Profitable Penalties for Lower Level Courts’, Judicature, Vol. 73, No. 5, 1990, pp. 248-252; T. Weigend, ‘Germany Reduces Use of Prison Sentences’, in M. Tonry & K. Hatlestad (Eds.), Sentencing Reforms in Overcrowded Times: A Comparative Perspective, Oxford University Press, New York and Oxford, 1997, pp. 177-178 (hereinafter: ‘Sentencing Reforms’); P.J.P. Tak, ‘Sentencing and Punishment in the Netherlands’, in Sentencing Reforms, pp. 194-200, at 196; A. Ashworth, ‘The Decline of English Sentencing and Other Stories’, in M. Tonry & R.S. Frase (Eds.), Sentencing and Sanctions in Western Countries, Oxford University Press, New York and Oxford, 2001, pp. 62-91, at 67 (hereinafter: ‘Sentencing and Sanctions’); T. Lappi-Seppälä, ‘Sentencing and Punishment in Finland: The Decline of the Repressive Ideal’, in Sentencing and Sanctions, pp. 92-150, at 92.

• 4 The Criminal Sanctions Agency, ‘Criminal Sanctions Agency Statements and Annual Report for the Year 2012’, 2013 (in Finnish), p. 4, available at <www.rikosseuraamus.fi/material/attachments/rise/julkaisut-monisteetjaraportit/6FcvDvctb/1_2013_TP_ja_toimintakertomus_2012_korj220313VALMIS.pdf>, accessed on 12 February 2014 (Finland – average); M. Nellis, K. Beyens & D. Kaminski (Eds.), Electronically Monitored Punishments: International and Critical Perspectives, Routledge, Abingdon, 2013, pp. 150-153, 179-182 (Netherlands and Belgium – low security prison); P. Lindström & E. Leijonram, ‘The Swedish Prison System’, in P.J.P. Tak & M. Jendly (Eds.), Prison Policy and Prisoners’ Rights, Proceedings of the Colloquium of the IPPF, Wolf Legal Publishers, Nijmegen, 2008, pp. 559-570, at 563 (hereinafter: “Prisons’ Rights”) (Sweden – average of low and high security prison); A. Barbarino & G. Mastrobuoni, ‘The Incapacitation Effect of Incarceration: Evidence from Several Italian Collective Pardons’, IZA Discussion Paper No. 6360, 2012, p. 33, available at <http://ssrn.com/abstract=2010975>, accessed on 28 January 2013 (Italy); Prison Reform Trust, Bromley Briefings Prison Factfile, Prison Reform Trust, London, 2013, p. 73 (England and Wales – average of 12 private prisons); J. Bungerfeldt, ‘Electronic Monitoring in Sweden: The Past, Present and the Future’, CEP Conference Stockholm, 2012, slide 12 (Sweden – low security); M. Mellett, ‘Irish Prison Policy, Prison Regime and Prisoners’ Rights’, in P.J.P. Tak & M. Jendly (Eds.), Prison Policy and Prisoners’ Rights, Proceedings of the Colloquium of the IPPF, Wolf Legal Publishers, Nijmegen, 2008, pp. 407-427, at 411 (Ireland – average); C. Henrichson & R. Delaney, The Price of Prisons What Incarceration Costs Taxpayers, Vera Institute for Justice, New York, 2012, p. 10, available at <www.vera.org/pubs/special/price-prisons-what-incarceration-costs-taxpayers>, accessed on 3 April 2014 (United States – average and range. The lowest costs are in Kentucky, and the highest are in New York State).

• 5 ‘Private prisons’ and ‘contracted prisons’ are used interchangeably in this article.

• 6 J. Dilulio, Jr., ‘What’s Wrong with Private Prisons’, Public Interest, Vol. 92, 1988, p. 66; C.H. Logan, Private Prisons: Cons and Pros, Oxford University Press, New York, 1990; R. Hardin, ‘Private Prisons’, Crime and Justice, Vol. 28, 2001, p. 265; D.E. Pozen, ‘Managing a Correctional Marketplace: Prison Privatization in the United States and the United Kingdom’, Journal of Law & Politics, Vol. XIX, 2003, p. 253.

• 7 K.L. Avio, ‘On Private Prisons: An Economic Analysis of the Model Contract and Model Statute for Private Incarceration’, New England Journal on Criminal & Civil Confinement, Vol. 17, 1991, p. 265; K.L. Avio, ‘The Economics of Prisons’, European Journal of Law and Economics, Vol. 6, 1998, p. 143; O. Hart, A. Shleifer & R.W. Vishny, ‘The Proper Scope of Government: Theory and an Application to Prisons’, The Quarterly Journal of Economics, Vol. 112, 1997, p. 1127

• 8 The notions ‘government’ and ‘state’ are used interchangeably in this article.

• 9 R.F. Culp, ‘The Rise and Stall of Prison Privatization: An Integration of Policy Analysis Perspectives’, Criminal Justice Policy Review, Vol. 16, 2005, p. 412.

• 10 To the best knowledge of the author, besides France, only Denmark, Prison Privatisation Report International, 2004, available at <www.psiru.org/justice/PPRI60.htm>, and one Land in Germany (Hessen) adopted to some extent the semi-private model of prison, at <www.serco.com/media/internationalnews/sercotorunfirstprivateprisoningermany.asp>, accessed on 18 March 2013.

• 11 D.C. McDonald, ‘Private Penal Institutions’, Crime and Justice, Vol. 16, 1992, p. 361; Ministry of Justice (UK), Contracted-Out Prisons, 2014, available at <www.justice.gov.uk/about/hmps/contracted-out>, accessed on 23 May 2014.

• 12 L.E. Glaze & E.J. Herberman, Correctional Populations in the United States, 2012, Bureau of Justice Statistics, Washington DC, 2013, p. 10, available at <www.bjs.gov/index.cfm?ty=pbdetail&iid=4843>, accessed on 19 March 2014.

• 13 The private company may own and operate the correctional institution, own and rent it to the state, or only provide prison services to publicly owned prisons, see McDonald 1992.

• 14 Hardin 2001.

• 15 L. Lanza-Kaduce, K.F. Parker & C.W. Thomas, ‘A Comparative Recidivism Analysis of Releases from Private and Public Prisons’, Crime & Delinquency, Vol. 45, 1999, p. 28.

• 16 E.C. Viano, ‘America’s Prison System’, in Prisoners’ Rights, pp. 139-184.

• 17 Hardin 2001.

• 18 Pozen 2003.

• 19 Logan 1990; Hardin 2001; S. Carbral & S. Saussier, ‘Organizing Prisons through Public-Private Partnership: A Cross-Country Investigation’, Brazilian Administration Review, Vol. 10, 2013, p. 100; B.W. Lundahl et al., ‘Prison Privatization: A Meta-Analysis of Cost and Quality of Confinement Indicators’, Research on Social Work Practice, Vol. 19, 2009, p. 383; T.C. Pratt & J. Maahs, ‘Are Private Prisons More Cost-Effective than Public Prisons? A Meta-Analysis of Evaluation Research Studies’, Crime & Delinquency, Vol. 45, 1999, p. 358.

• 20 Lanza-Kaduce et al. 1999; W.D. Bales et al., ‘Recidivism of Public and Private State Prison Inmates in Florida’, Criminology & Public Policy, Vol. 4, 2005, p. 57; A.L. Spivak & S.F. Sharp, ‘Inmate Recidivism as a Measure of Private Prison Performance’, Crime & Delinquency, Vol. 54, 2008, p. 482.

• 21 N. Panchamia, Competition in Prisons, Institute for Government, London, 2012, p. 1.

• 22 Ministry of Justice 2014.

• 23 Prison Reform Trust, ‘Bromley Briefings Prison Factfile’, 2013, p. 72.

• 24 See e.g. Prison Reform Trust, ‘Private Punishment: Who Profits?’, 2005.

• 25 Ibid.; Pozen 2003.

• 26 Hardin 2001.

• 27 Pozen 2003.

• 28 Hardin 2001; Pozen 2003. In addition, England and Wales piloted with an alternative payment scheme, i.e. the PbR model. For an explanation, see Section 3.3.3.

• 29 G. Sturgess, ‘The Sources of Benefit in Prison Contracting’, in V. Helyar-Cardwell (Ed.), Delivering Justice: The Role of the Public, Private and Voluntary Sectors in Prisons and Probation, Criminal Justice Alliance, London, 2012, pp. 31-40; W. Tanner, ‘The Case for Private Prisons’, Reform Ideas, No. 2, 2013; Panchamia 2012.

• 30 U. Rosenthal & B. Hoogenboom, ‘Some Fundamental Questions on Privatisation and Commercialisation of Crime Control, with Special Reference to Developments in the Netherlands’, in Privatisation of Crime Control: Collected Studies in Criminological Research (Vol. XXVII), Council of Europe, Strasbourg, 1990, pp. 17-43.

• 31 Carbral & Saussier 2013.

• 32 For the explanation of the ‘cost-plus contract’, see P. Bajari & S. Tadelis, ‘Incentives versus Transaction Costs: A Theory of Procurement Contracts’, RAND Journal of Economics, Vol. 32, 2001, pp. 387-388.

• 33 Carbral & Saussier 2013.

• 34 Cour des Comptes, Garde et Reinsertion: La Gestion des Prisons, Rapport Public Thématique, 2006, p. 175.

• 35 As defined in the introduction. Namely, contracted-out prisons sponsored by the state, rather than a punishment entirely supplied by the private market.

• 36 A. Ogus, Regulation: Legal Form and Economic Theory, Hart Publishing, Portland, 2004.

• 37 R. Cooter & T. Ulen, Law and Economics (4th edn), Addison Wesley, Boston, 2003.

• 38 Ogus 2004.

• 39 Cooter & Ulen 2003.

• 40 Other goals might be to provide rehabilitation for offenders. Nevertheless, regardless of the methods used in different prisons to deal with prisoners, the end of all custodial sanctions is to protect the society from crime. This might be achieved through incapacitation, deterrence, rehabilitation, and so on.

• 41 W.A. Niskanen, ‘Bureaucrats and Politicians’, Journal of Law and Economics, Vol. 18, 1975, p. 617; Ogus 2004; M. Boycko, A. Shleifer & R.W. Vishny, ‘A Theory of Privatisation’, The Economic Journal, Vol. 106, 1996, p. 309.

• 42 Boycko et al., Ibid.

• 43 A. Shleifer, ‘State versus Private Ownership’, The Journal of Economic Perspectives, Vol. 12, 1998, pp. 133-150.

• 44 Ogus 2004, p. 277.

• 45 Shleifer 1998.

• 46 S.O. Richardson, The Political Economy of Bureaucracy, Routledge, Abingdon, 2011; Logan 1990.

• 47 If politicians are hiring based on loyalty, every politician would be interested in promoting his own supporters, thus replacing the old managers.

• 48 Hart et al. 1997; Sturgess 2012.

• 49 Tanner 2013.

• 50 G. De Geest, ‘Who Should Be Immune from Tort Liability?’, Journal of Legal Studies, Vol. 41, 2012, p. 291. The example provided by the author is that of fire fighters. A fire fighter needs to balance between the damage which may be caused due to excessive usage of water on the one hand and a fire damage of the property due to insufficient action on the other hand. In a situation where the fire fighter may be held liable for the water damage, avoiding this measure is a precaution. Since the precaution costs are mainly externalized to the society (a burning house), while the harm is internalized by the fire fighter, he is expected to choose to overinvest in precaution costs, i.e. to avoid action. This result explains why some officials have immunity from liability, e.g. police officers (p. 292).

• 51 For example, following an escape of a notorious child-molester and a murderer Marc Dutroux, two Belgian ministers (the Minister of Interior and the Minister of Justice), along with the police chief, had resigned. See <www.bendevannijvel.com/andere/dutroux_inleiding.html>, accessed on 12 May 2014 (in Dutch). Another example may be found in Sweden where following the escape of a Swedish spy, the Justice Minister had resigned. See <www.nytimes.com/1987/10/20/world/a-swede-resigns-over-spy-s-escape.html>, accessed on 12 May 2014. See also Logan 1990, p. 71, asserting that public officials are more responsive to political loses rather than economic losses since the later can be externalized to the taxpayers. As a result, politicians will overspend on security.

• 52 N.G. Poythress & S.L. Brodsky, ‘In the Wake of a Negligent Release Law Suit: An Investigation of Professional Consequences and Institutional Impact on a State Psychiatric Hospital,’ Law and Human Behavior, Vol. 16, 1992, p. 155.

• 53 For example, in one maximum-security prison in the Netherlands, there are currently five prisoners and around 60 guards. This ratio seems excessive and inefficient. See H.G. van de Bunt et al., Gevangen in de EBI. Een empirisch onderzoek naar de Extra Beveiligde Inrichting (EBI) in Vught, Boom Lemma Uitgevers, Den Haag, 2013.

• 54 A. Downs, Inside Bureaucracy, Little, Brown and Company, Boston, 1967, pp. 25-30.

• 55 W.A. Niskanen, Jr., Bureaucracy and Representative Government, Aldine Atherton, Chicago, 1971, pp. 38-39.

• 56 Richardson 1990, pp. 25-26; Downs 1967, p. 196.

• 57 Ogus 2004, pp. 277-278.

• 58 Ibid., p. 286.

• 59 J. Kornai, E. Maskin & G. Roland, ‘Understanding the Soft Budget Constraint’, Journal of Economic Literature, Vol. 41, 2003, p. 1095; Logan 1990.

• 60 Ogus 2004, pp. 272-273.

• 61 G. Tullock, The Selected Works of Gordon Tullock: Bureaucracy, Vol. 6, Library Fund, Inc., Indianapolis, 2005, p. 283.

• 62 A. Shleifer, ‘Theory of Yardstick Competition’, The RAND Journal of Economics, Vol. 16, 1985, p. 319.

• 63 Hardin 2001, pp. 333-334; Tanner 2013, pp. 12-13; Panchamia 2012, p. 5; Sturgess 2012, pp. 33-40.

• 64 Downs 1967, pp. 160, 195.

• 65 Hart et al. 1997, p. 1129.

• 66 Downs 1967, p. 59.

• 67 Avio 1998, p. 150 (emphasis added).

• 68 For the application of other economic models to prison privatization, see e.g. Avio 1998, using the ‘product quality model’; Hart et al. 1997, applying the ‘incomplete contract model’.

• 69 M.C. Jensen & W.H. Meckling, ‘Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure’, Journal of Financial Economics, Vol. 3, 1976, pp. 305, 308, 313.

• 70 As long as the marginal costs of crime prevention does not exceed the marginal benefits derived from it. See G.S. Becker, ‘Crime and Punishment: An Economic Approach’, Journal of Political Economy, Vol. 76, 1968, pp. 169, 207-209.

• 71 For the discussion on the lobby argument and evidence for its limited practice, see Hardin 2001, pp. 278-283.

• 72 Hardin 2001, p. 289.

• 73 Lanza-Kaduce et al. 1999, p. 30.

• 74 Hardin 2001, p. 307.

• 75 Some evidence of successful privatization in different fields may be found in the World Bank, Bureaucrats in Business: The Economics and Politics of Government Ownership, The International Bank for Reconstruction, Washington DC, 1995.

• 76 It seems that the prescriptive character of the British contracts indeed minimized innovation. Panchamia 2012. In 2011, the Ministry of Justice relaxed the strictness of the contracts in order to leave more space for the providers to decide on the methods to deliver their services.

• 77 K.J. Murphy, ‘Executive Compensation’, in R. Layard, O. Ashenfelter & D.E. Card (Eds.), Handbook of Labor Economics (1st edn, Vol. 3, Chapter 38), Elsevier, 1999, pp. 2485, 2519, 2527, 2531-2532.

• 78 Avio 1991, pp. 285-286.

• 79 See <www.justice.gov.uk/contacts/prison-finder/doncaster>, accessed on 23 March 2014.

• 80 M. Chambers, ‘Expanding Payment-by-Results: Strategic Choices and Recommendations’, Policy Exchange, Vol. 1, 2012, pp. 11-12. The PbR was also tested in the public sector and with other services. Other piloted models were paying bonus on top of cost coverage for reducing recidivism; or paying no income unless the expected results are met (p. 11).

• 81 An official response from the UK Ministry of Justice to the author’s inquiry (11 April 2014).

• 82 J.E. Core, W.R. Guay & D.F. Larcker, ‘Executive Equity Compensation and Incentives: A Survey’, Economic Policy Review, Vol. 9, 2003, pp. 27, 30.

• 83 A. Blumstein et al. (Eds.), Criminal Careers and “Career Criminals”, National Academy Press, Washington DC, 1986, p. vii; In 1994, around 70% of prisoners in the United States were rearrested within three years after their release. Around 52% returned to prison during this period, see T. Hughes & D.J. Wilson, Reentry Trends in the United States, Bureau of Justice Statistics, 2002, pp. 20-21.

• 84 For instance, S.D. Levitt, ‘The Effect of Prison Population Size on Crime Rates: Evidence from Prison Overcrowding Litigation’, The Quarterly Journal of Economics, Vol. 111, 1996, p. 319; K.E. McCollister, M.T. French & H. Fang, ‘The Cost of Crime to Society: New Crime-Specific Estimates for Policy and Program Evaluation’, Drug and Alcohol Dependence, Vol. 108, 2010, p. 98.

• 85 Core et al. 2003, p. 32.

• 86 Lanza-Kaduce et al. 1999, p. 34.

• 87 It does not seem reasonable to believe that the private providers may invent a method which reduces criminal inclination only for a limited period, after which the offender would always return to crime. No doubt, in practice this might occur if, e.g. the legal opportunities of the released offender change after years. In this case, the benefits of crime may once again overweight its costs. However, the private prisons have no control over such situations.

• 88 Carbral & Saussier 2013, p. 107.

• 89 Hardin 2001; Lundahl et al. 2009.

• 90 Tanner 2013, pp. 15-17.

• 91 Prison Reform Trust 2005, p. 8.

• 92 For example, the high-security prison in the Netherlands currently accommodates five prisoners and around 60 members of the staff. The high number of the prison employees may not even be justified by the reasoning that they are required in order to prevent violence of the prisoners. The internal guidelines of this prison state that the prison officers are not allowed to face more than one prisoner at the same time. Thus, in case of a violent event of more than one prisoner, the staff is prohibited from intervening. The costs of such a prison seem exceeded and doubtful if the number of personnel may be considered optimal (van de Bunt 2013).

• 93 Sturgess 2012, p. 34.

• 94 A counter argument may be that separating the ownership from management might improve the accountability. In the United Kingdom, for instance, through the Market Testing procedure, the management of the prisons may be returned to the hands of the public sector if it outperforms the private provider. This is hard to do if the prison is owned by the private party and the buyback period is long, e.g. 20-40 years in the United Kingdom (Hardin 2001, p. 323). A possible way to resolve this issue is by requiring the private firms to agree in advance for a rent contract in case they underperform, and the management is transferred to the public sector. This provision in the contract may in fact add incentives to perform better than the public sector. Another solution is to provide a contract for construction and management but to maintain the ownership in the hands of the state. This way the private provider is still incentivized to produce an innovative design, yet there is no impediment for the Market Testing since the facility is owned by the state.

• 95 M. Bozovic (Ed.), Jeremy Bentham, The Panopticon Writings, Verso, London, UK, 1995. The letters were written in the 18th century.

• 96 Not overcrowded prisons, proper nutrition, good health care, and so on.

• 97 It might be expected that repeating negative coverage regarding violation of human rights in private prisons would lead to a public pressure to cancel the particular contract or to abolish prison privatization in general.

• 98 Pozen 2003, pp. 277-280.

• 99 Logan 1990, p. 204.

• 100 Dilulio 1988, pp. 79, 81; S. McConville, ‘The Privatisation of Penal Services’, in Privatisation of Crime Control, Collected Studies in Criminological Research (Vol. XXVII), Council of Europe, Strasburg, 1990, pp. 77, 94.

• 101 Hardin 2001, p. 278.

• 102 The Constitution of Finland, 11 June 1999 (731/1999, amendments up to 1112/2011 included).

• 103 T. Lappi-Seppälä, ‘Imprisonment and Penal Policy in Finland’, in P. Wahlgren (Ed.), Scandinavian Studies in Law, Vol. 54, Stockholm Institute for Scandinavian Law, Stockholm, 2009, pp. 333, 344; See also McConville 1990, p. 83, stating that even in jurisdictions where there is no explicit prohibition in the constitution for private administration of penal institutions, it might be implied.

• 104 Hardin 2001, p. 266.

• 105 Richardson 1990, p. 27.

• 106 Logan 1990, p. 54.

• 107 See e.g. escapes and sexual abuse by public workers in preventive detention (TBS) in the Netherlands, available at <www.sevendays.nl/artikel/124986>, <www.misbruikdoorhulpverleners.nl/tbsgog&tbs.html>, accessed on 21 May 2014 (in Dutch); overcrowding and mal conditions in Italian prisons, available at <http://espresso.repubblica.it/inchieste/2014/04/25/news/carceri-ecco-perche-siamo-la-vergogna-d-europa-1.162997>, accessed on 21 May 2014 (in Italian).

• 108 Logan 1990, pp. 55, 75.

• 109 Shleifer 1998, pp. 16-17.

• 110 Logan 1990, pp. 11-12.

• 111 See e.g. the objection of the European Public Service Union (EPSU) to prison privatization in Denmark and in the Netherlands. One of their explicit arguments for the objection is that the staff would receive lower salaries and lesser benefits. Available at <www.mijnvakbond.nl/Waarschuwing-tegen-privatisering-gevangenissen?referrer=1133> (in Dutch) and at <www.epsu.org/a/7376>, accessed on 12 May 2014; similar resistance by the trade union to private prisons may be found in Belgium, available at <www.gva.be/archief/guid/acod-categoriek-tegen-privatisering-gevangeniswezen.aspx?artikel=acfadfd4-df43-4fd3-a136-9a8d58b8836a>, accessed on 12 May 2014 (in Dutch); in France, the National Union of Prison Directors (SNDP) also opposed contracting out prison services to private parties, available at <www.liberation.fr/societe/2012/01/31/prisons-gare-a-la-privatisation-du-service-public-penitentiaire_792402>, accessed on 12 May 2014 (in French).