influence of regulation on the productivity of waste utilities. what can we learn with the...

10
Influence of regulation on the productivity of waste utilities. What can we learn with the Portuguese experience? Pedro Simões , Rui Cunha Marques CEG-IST, Technical University of Lisbon, Avenida Rovisco Pais, 1049-001 Lisbon, Portugal article info Article history: Received 5 September 2011 Accepted 5 February 2012 Available online 2 March 2012 Keywords: Productivity Rate of return regulation Regulation Sunshine regulation Urban waste sector abstract This paper examines the merits and the perverse effects of quality of service regulation in the performance of urban waste services when implemented alone and compares the performance of differ- ent economic regulatory methods. By means of a productivity analysis, we investigate the influence of a five-year period of regulation on the performance of Portuguese urban waste utilities using an unbal- anced panel data for the period 2001–2008. Different non-parametric methods were applied to estimate the productivity change, all leading to similar outcomes. We observed a tendency of productivity decline in the urban waste utilities and concluded that in spite of the unequivocal improvements in the quality of service induced by sunshine regulation, more should be done as far as economic regulation is concerned. We also found that the use of sunshine regulation together with low incentive economic regulatory methods is not positive, leading to overinvestment rather than to value for money. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction The Portuguese waste sector has undergone several reforms in the past two decades (Pássaro, 2003). In particular, the implemen- tation of a sector-specific regulator for the waste services is one of the most notable ones. The creation of this regulator was a re- sponse to the unfavourable context that characterised the waste sector in the past. Some of the features of that period were related to a growing tendency for waste production, the need to divert waste from landfills, the presence of inherent market distortions, the existence of tariffs far below the actual costs of service, and the weight that expenses gained in the State Budget. In this regard, regulation might be very important for the protection of the public interest. The particular features of the Portuguese regulatory model have several consequences in its performance. One of them is that only a part of the ‘‘wholesale’’ market (sorting, transport, treatment and disposal) is regulated (only the concessionaire companies) and the remaining (non-concessionaire), either public or private, is not regulated. The retail market is not regulated at all (since there are no concession contracts). The two most noteworthy issues of this regulatory model are related to the quality of service regula- tion and the method used for economic regulation. For the quality of service, the Portuguese waste regulator, the ERSAR (The Water and Waste Services Regulation Authority), uses the sunshine regu- latory model (based on comparison and public discussion of per- formance) which has led to good results (Marques and Simões, 2008). The rate-of-return regulation (RoR) is adopted for economic regulation but this method does not encourage the utilities to be- come more efficient and productive (Aubert and Reynaud, 2005). In opposition to performance based regulation (e.g. price cap regula- tion) where price limits are defined and the companies receive incentives since they retain the profits of their outperformance, in RoR the payments are based on costs and the profits will be the same whether the performance is good or bad. We suspect that the mixture between RoR and sunshine regulation is a game of negative sum, since the quality of service improvement requires investments and extra costs and the RoR stimulates overinvest- ment (Averch-Johnson effect). Then, the quality of service enhance- ment can result in prices increase (borne by customers and/or taxpayers whose willingness to pay has not been assessed), distort- ing the potential positive benefits of regulation. Furthermore, usu- ally there is a close relationship between operators and construction firms and the politicians are always in favour of investments in new public infrastructures. As such, the entire con- text stimulates investments (both good and bad) which can result in gold-plating practices or lack of ‘‘true’’ incentives provided to the market (Jamison and Berg, 2004). In these circumstances, reg- ulation will not provide the right value for money. Although recent, regulation in the waste sector is a matter with increasing interest for academics and practitioners. Several coun- tries are creating their regulatory authorities (Italy, Romania, Brazil, Kosovo, etc.). As waste services are starting to be paid by users in many countries and are endowed with several market 0956-053X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2012.02.004 Corresponding author. E-mail addresses: [email protected] (P. Simões), [email protected] (R.C. Marques). Waste Management 32 (2012) 1266–1275 Contents lists available at SciVerse ScienceDirect Waste Management journal homepage: www.elsevier.com/locate/wasman

Upload: pedro-simoes

Post on 30-Oct-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Waste Management 32 (2012) 1266–1275

Contents lists available at SciVerse ScienceDirect

Waste Management

journal homepage: www.elsevier .com/ locate/wasman

Influence of regulation on the productivity of waste utilities. What can welearn with the Portuguese experience?

Pedro Simões ⇑, Rui Cunha MarquesCEG-IST, Technical University of Lisbon, Avenida Rovisco Pais, 1049-001 Lisbon, Portugal

a r t i c l e i n f o

Article history:Received 5 September 2011Accepted 5 February 2012Available online 2 March 2012

Keywords:ProductivityRate of return regulationRegulationSunshine regulationUrban waste sector

0956-053X/$ - see front matter � 2012 Elsevier Ltd.doi:10.1016/j.wasman.2012.02.004

⇑ Corresponding author.E-mail addresses: [email protected] (P. Sim

(R.C. Marques).

a b s t r a c t

This paper examines the merits and the perverse effects of quality of service regulation in theperformance of urban waste services when implemented alone and compares the performance of differ-ent economic regulatory methods. By means of a productivity analysis, we investigate the influence of afive-year period of regulation on the performance of Portuguese urban waste utilities using an unbal-anced panel data for the period 2001–2008. Different non-parametric methods were applied to estimatethe productivity change, all leading to similar outcomes. We observed a tendency of productivity declinein the urban waste utilities and concluded that in spite of the unequivocal improvements in the quality ofservice induced by sunshine regulation, more should be done as far as economic regulation is concerned.We also found that the use of sunshine regulation together with low incentive economic regulatorymethods is not positive, leading to overinvestment rather than to value for money.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The Portuguese waste sector has undergone several reforms inthe past two decades (Pássaro, 2003). In particular, the implemen-tation of a sector-specific regulator for the waste services is one ofthe most notable ones. The creation of this regulator was a re-sponse to the unfavourable context that characterised the wastesector in the past. Some of the features of that period were relatedto a growing tendency for waste production, the need to divertwaste from landfills, the presence of inherent market distortions,the existence of tariffs far below the actual costs of service, andthe weight that expenses gained in the State Budget. In this regard,regulation might be very important for the protection of the publicinterest.

The particular features of the Portuguese regulatory model haveseveral consequences in its performance. One of them is that only apart of the ‘‘wholesale’’ market (sorting, transport, treatment anddisposal) is regulated (only the concessionaire companies) andthe remaining (non-concessionaire), either public or private, isnot regulated. The retail market is not regulated at all (since thereare no concession contracts). The two most noteworthy issues ofthis regulatory model are related to the quality of service regula-tion and the method used for economic regulation. For the qualityof service, the Portuguese waste regulator, the ERSAR (The Waterand Waste Services Regulation Authority), uses the sunshine regu-

All rights reserved.

ões), [email protected]

latory model (based on comparison and public discussion of per-formance) which has led to good results (Marques and Simões,2008). The rate-of-return regulation (RoR) is adopted for economicregulation but this method does not encourage the utilities to be-come more efficient and productive (Aubert and Reynaud, 2005). Inopposition to performance based regulation (e.g. price cap regula-tion) where price limits are defined and the companies receiveincentives since they retain the profits of their outperformance,in RoR the payments are based on costs and the profits will bethe same whether the performance is good or bad. We suspect thatthe mixture between RoR and sunshine regulation is a game ofnegative sum, since the quality of service improvement requiresinvestments and extra costs and the RoR stimulates overinvest-ment (Averch-Johnson effect). Then, the quality of service enhance-ment can result in prices increase (borne by customers and/ortaxpayers whose willingness to pay has not been assessed), distort-ing the potential positive benefits of regulation. Furthermore, usu-ally there is a close relationship between operators andconstruction firms and the politicians are always in favour ofinvestments in new public infrastructures. As such, the entire con-text stimulates investments (both good and bad) which can resultin gold-plating practices or lack of ‘‘true’’ incentives provided tothe market (Jamison and Berg, 2004). In these circumstances, reg-ulation will not provide the right value for money.

Although recent, regulation in the waste sector is a matter withincreasing interest for academics and practitioners. Several coun-tries are creating their regulatory authorities (Italy, Romania,Brazil, Kosovo, etc.). As waste services are starting to be paid byusers in many countries and are endowed with several market

P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275 1267

failures, a ‘visible hand’ (a regulator) is required to correct andmitigate them. Therefore, the aim of this article is to investigatethe influence of regulation on the performance of waste utilities.For this purpose, we intend to test several hypotheses, respondingto the research questions below while contributing to theliterature:

1.1. Has the ‘sunshine regulation’ method produced positive results inthe quality of service regulation?

The sunshine regulation of the Portuguese waste sector relieson the comparison, discussion and public display of the perfor-mance of a public service by conducting reports, studies, opinionsand analysis (Braconnier, 2001). In this model, the regulator has nopower to impose serious penalties and its power relies on thethreat that is felt by the utilities because their defects or deficien-cies are exposed to the general public by the regulator (Marques,2006). This ‘name and shaming’ strategy is considered as very po-sitive, mainly in the quality of service regulation, since its improve-ment does not necessarily correspond to a reduction of revenuesand profits. Here we try to demonstrate whether this hypothesisis true or not.

1.2. Does the ‘sunshine regulation ’method have a perverse effect onthe utility productivity?

Despite the good results pointed out as far as quality of serviceis concerned in Portugal, until now it has not been clear if there arebenefits in terms of productivity improvement of the service pro-vided. In principle, these results cannot be accompanied by thesame kind of success in economic terms (Spasovic et al., 1999).The regulated waste utilities might show low productivity andtransfer the cost of the increase in the quality of service perfor-mance to the tariffs paid by the customers (or taxpayers). We eval-uate this hypothesis for the Portuguese case-study.

1.3. Is the low productivity change compensated by the improvementin the quality of service?

If productivity declined in the period analysed (correspondingto the existence of regulation), this could have been a result ofthe improvement in the quality of service. Although this mightbe a matter of customer willingness to pay for the service, weinvestigated if there was some kind of counterweight between pro-ductivity and quality of service.

1.4. Does RoR have a negative effect in the productivity change,compared with the other economic regulatory methods?

RoR defines a priori a fixed profit for the regulated entities. It al-lows for and guarantees a fair and ‘reasonable’ (fixed) profit for theregulated utilities, irrespective of their actual results and perfor-mance. It is characterised by a short regulatory period and a lowrisk profile, but encourages the steadiness and ‘quiet-life’ of man-agers. Additionally, it fosters overinvestment (the so-calledAverch-Johnson effect) and sometimes gold-plate practices. Con-cerning this hypothesis, we analyse the influence of RoR on theutilities productivity. As most of the non-regulated companies(mostly private ones) are subject to a regulation by contract modelwhere the prices (and the updating formula) are set at the begin-ning, in practical terms they are subject to a price cap regimeand therefore we can compare different economic regulatoryregimes.

1.5. Is the mixture of RoR and sunshine regulation risky in terms of thescarce incentives provided and value for money provided?

As sunshine regulation and RoR seem to stimulate investmentand additional expenses, this circumstance might be penalisingfor the productivity of waste utilities and consequently for custom-ers or taxpayers. Besides, incentives are not provided with RoRsince the ‘rate of return’ of waste utilities is the same, whether theyshow the best or the worst practices (Kahn, 1988). As gold-platingpractices sometimes are welcomed in public utilities (both by cus-tomers and politicians) the value for money might indeed bejeopardised. This issue will be investigated for the Portuguesecase-study.

To test these hypotheses and demystify these so important top-ics, we applied different approaches to measure the productivity,and to provide robustness to the results. In this analysis, we usean eight-year sample for all the ‘‘wholesale’’ waste market (themost regulated segment), covering 5 years of ERSAR’s regulatoryactivity. In addition, the analysis encompasses all the Portuguesenon-regulated companies (in the ‘wholesale’ market) in that period(the non-concessionaire companies), allowing for the performancecomparison with the regulated ones. Despite not being new inother infrastructure services, such as the energy or railways, theinclusion of a quality index in the productivity analysis in thewaste sector is definitely an interesting feature (Färe et al., 1995)and a sound contribution to the literature. This procedure intendsto merge in the analysis the obligations in quality of serviceimprovement and the economic (regulatory) incentives. Beyondthe goal of evaluating the influence of RoR and sunshine regulationtogether and the possibility of overinvestment and extra costs, thepaper tries to assess if, at some point, the quality of serviceimprovement may counterweigh the productivity results.

After this brief introduction, the remainder of this article isstructured as follows. Section 2 reviews the literature of other sec-tors regarding the influence of regulation on their performance.Section 3 presents the regulatory model adopted in Portugal andthe recent reforms carried out. Section 4 introduces the methodol-ogies used for productivity measurement. Section 5 computes thedifferent models that allow for testing the mentioned hypotheses.Section 6 discusses the results and provides some policy implica-tions and, finally, Section 7 concludes the paper.

2. Overview of the literature on the influence of regulation

The influence of regulation on infrastructure services has beenanalysed by means of different econometric and mathematical pro-gramming approaches. However, the scope of these studies widelydiverges among the sectors. Despite the clear absence of studiescovering the productivity analysis of the waste sector, there issome research documenting the performance evaluation (e.g. San-chez, 2008; Benito et al., 2010) and the effects of the operationalenvironment and, fundamentally, the effect of privatisation onthe utilities’ performance (Bosch et al., 2000; Bel and Costas,2006 , in Spain, Ohlsson, 2003, in Sweden or Dijkgraaf and Gradus,2003, in the Netherlands), the effect of competitive tendering(Domberger et al., 1986; Gómez-Lobo and Szymanski, 2001.) orthe effect of the implementation of local policies (Worthingtonand Dollery, 2001; De Jaeger et al., 2011). The analysis of the influ-ence of regulation on productivity is much more limited. In fact,until 2010, Chapple and Harris (2003) conducted the single workthat, to our knowledge, exists on this matter. They evaluated theinfluence of regulation (legislation) and a landfill tax on waste pro-duction by means of a Malmquist-Luenberger productivity index.This study reports to the period between 1991 and 1998 andencompasses 42 English counties/metropolitan areas.

1268 P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275

When the scope is extended to other utility services, the sce-nario is completely different. Besides the vast number of studiesthat analyses the productivity change, the literature on regulationis relevant and there is a handful of studies including influence ofquality on frontier analysis, even though they are not related to thesector specific regulation.

Regarding the telecoms sector, conclusions about the differenttypes of regulation have been mixed. For instance, there is no con-sensus about the effects of the different models of regulation, espe-cially when the quality of service is considered. Uri (2002)estimated positive effects of incentive regulation in the US tele-communications (19 local exchange carriers, between 1988 and1999). The same conclusion was reached by Lam and Lam(2005), analysing the dominant firm in Hong Kong’s local tele-phone market between 1964 and 1998, and by Seo and Shin(2011) also in the US telecoms industry (25 LECs 1988–1998). Aiand Sappington (2002) concluded that costs were lower underprofit sharing and price cap regulation when local competitionwas sufficiently intense in the US telecommunications industry be-tween 1986 and 1999. Roycroft (1999) found benefits from pricecap and incentive regulation in productivity of Ameritech operat-ing companies. Moreover, Eckenrod (1999) observed that the aver-age price markup increased slightly after price cap regulation hasbeen implemented, with some benefits in the long-term. In oppo-sition, Orr and Lefebvre (1993) evaluated the price performance inthe Canadian (regulated monopoly) and in the US (liberal long-dis-tance market entry policy) telecom industry. They found that reg-ulated competition had not improved performance. When qualityof service was considered, Resende and Facanha (2005) showedpoor quality performance under price-cap regulation in the UStelecoms sector.

For the electricity sector, Ter-Martirosyan (2003) found thatincentive regulation reduced the US utility’s expenses (1993–1999). However, this is also associated with deterioration in somequality indicators. Ida et al. (2007) identified positive effects of theregulatory reforms on the Japanese electric power companies be-tween 1996 and 2002. Fleishman et al. (2009) analysed the effectof regulation (namely on air quality) in more than 2000 US powerplants from 1994 to 2004 and identified mixed effects of regula-tion. Giannakis et al. (2005) also found a trade-off between effi-ciency and quality for the electricity distribution in the UK,between 1991/92 and 1998/99, using a quality-incorporated MPI.Estache and Rossi (2005) concluded that incentive-based regimeslead to higher labour productivity than RoR in Latin America.Nakano and Managi (2008) examined the effect of reforms onJapan’s steam power-generation over the period 1978–2003 andobserved that the regulatory reforms have contributed to a produc-tivity growth. Pérez-Reyes and Tovar (2009) examined the influ-ence of the reforms on 14 electricity distribution companies inPeru between 1996 and 2006 and some improvements were ob-served. Zhang et al. (2008) using panel data for 36 developingand transitional countries, over the period 1985–2003, found thatprivatisation and regulation did not lead to obvious gains in eco-nomic performance. Cambini and Rondi (2009) investigated therelationship between investment and regulatory regimes (incen-tive vs. RoR) for a sample of EU energy utilities from 1997 to2007, concluding for a higher investment under incentive regula-tion. Berg et al. (2005) analysed 24 Ukraine electricity distributioncompanies from 1998 to 2002, and identified some benefits fromregulatory incentives. At last, Sanyal (2007) analysing the USelectric utilities from 1990 to 2001, found that deregulation hada substantial negative impact on expenditures.

Considering the water utilities, Saal and Parker (2001) con-cluded that between 1985 and 1999 privatisation and the pricecaps in the UK had a relevant influence on the reduction of thequality of service measured. Maziotis et al. (2009) evaluated the

effect of OFWAT price cap regulation in England and Wales. Theystated that throughout the entire 1991–2008 period price capswere never ‘‘powerful’’, in the sense that they required less produc-tive firms to immediately and fully catch-up with the most produc-tive firms to regain economic profitability. Aubert and Reynaud(2005) showed that the regulatory model (price-cap or RoR) canexplain the performance of the 211 water utilities in Wisconsin.Witte and Saal (2010) evaluated the effect of diverse reforms andmergers of companies and the positive influence of the bench-marking exercise of VEWIN (Dutch Water Association) for the timeperiod between 1992 and 2006. Lin and Berg (2008) found that thequality of service slightly declined between 1998 and 2001 in thewater industry in Peru by applying the quality-incorporated MPI.Motta and Moreira (2006) noticed that the lack of regulation inthe Brazilian sanitation sector did not stimulate advances in thetechnological frontier (1997–2002).

Finally, in transportation, Estache et al. (2007) studied thedichotomy between quantity and quality in the freight railwayscase in Brazil (1992–2001). Using the quality-incorporated MPI,they found the prevalence of quantity–quality trade-offs up tothe end of the reform period and a positive correlation betweenthem. Odeck (2008) revealed the improvements in terms ofproductivity with mergers in the Norwegian bus industry(1995–2002). Minken and Killi (2000) evaluated the influence ofmajor reforms in the Norwegian ferry links (113 ferry links on1988–1996) and the possible improvements with the setting ofquality standards. Santos et al. (2010) tried to understand the Por-tuguese reform in the railway sector, where losses were detected,even under a price-cap regulation regime since 2005. In the rail-road industry, Wilson (1997) found (1978–1989) that initially costsavings from partial deregulation were modest. However, by 1989,cost savings were tremendous, about 40 percent lower than theywould have been under regulation. Mizutani et al. (2009) foundbenefits from the yardstick competition in the Japanese rail com-panies (1995–2000). Margari et al. (2007) emphasised the impactof regulatory and environmental norms on public transit systems.Hammond et al. (2002) observed that firms operating under thebasic price system, which provides stronger incentives to reduceprices through efficiency savings, are shown to be more efficientthan those under the maximum price system. Benfratello(2009) studied 20 Italian motorways concessionaires over the1992–2004 period and found that their productivity did not in-crease with the adoption of a price cap regime.

To sum up, important questions have been raised in the litera-ture regarding the different (infrastructure) sectors. For instance,the apparent quality of service deterioration in the services pro-vided is, in many cases, related to too demanding incentives gener-ated by regulation of the market. However, the incentives arisingfrom RoR, when quality of service improvements are imposed,are often put in question in terms of productivity change andencourage overinvestment and excessive costs. In economic regu-lation, performance based regulation (e.g. price cap and revenuecap regulation) has proven its benefits, although quality of serviceand public service obligations should be assured and carefullysupervised. Nevertheless, the measures of quality are the maincause of disagreement among authors and are key aspects in anyregulated sector, mainly in essential public services as those re-lated to waste.

3. Regulation of the Portuguese waste sector

3.1. Economic regulation

In Portugal, the procedure to set waste tariffs depends on themanagement model of each waste utility. ERSAR has had responsi-

P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275 1269

bility in this issue only regarding concessionary companies andparticularly in the regional (‘wholesale’) waste systems. Based ontheir budgets and investment plans and on the annual account re-ports of the companies, it approves their budgets and their pro-posal for future tariffs. The formula for defining the tariff systemconsists of a hybrid methodology, based on the RoR method, estab-lished contractually, and introducing a mechanism to share gainsin productivity. Although it varies from contract to contract, therate of return is fixed by the concession granting authority andconsists of a base rate (without risk) plus a risk premium of 3%, ap-plied to the capital stock and the legal reserve. The ERSAR proposesthe tariffs which are approved by the Minister of Environment,who is simultaneously the ‘‘owner’’ of the ‘wholesale’ regionalcompanies (multimunicipal systems). Actually, the companiespresent their ‘generous and inefficient’ costs and the regulator,after some negotiation (with a little reduction in costs), proposesa tariff in line with the budget of the companies.

3.2. Quality of service regulation

In terms of quality of service regulation, the approach adoptedby ERSAR relies on the discussion, public display and regular com-parison of a set of performance indicators. This strategy, known assunshine regulation, has become a powerful and effective tool toprovide performance incentives by fostering (virtual) competitionbetween operators (Marques, 2006).

The performance evaluation of urban waste services adopted byERSAR encompasses a set of 20 performance indicators intendingto represent the most important aspects that may reflect the qual-ity of service. This set includes three distinct groups of indicators:(a) the defence of the user interests, (b) the operator sustainabilityand (c) the environmental sustainability. The regulator computesthese indicators since 2004. Table 1 shows the performance indica-tors used for the urban waste utilities for the years 2004 and 2009and the positive progress.

This regulatory model evaluates individually the performanceof each operator, comparing its results with reference values. Thelatter are target values that ERSAR deems as likely to be attainedby the utilities, based on historical values of each utility. ERSAR

Table 1Performance indicators for waste services in Portugal.

2004 2009

Protection of the users’ interestsService coverage (%) 100 100Selective collection coverage (%) 67 82a

Average service charges (€/ton) 26.6 25.0Answers to written complaints (%) 41 97

Operator sustainabilityOperating cost coverage ratio (�) 1.5 1.56Unit running costs (€/ton) 21.6 28.2Solvency ratio (�) 0.19 0.64Recycling (%) 4.1 8.5Organic recovery (%) 2.8 7.0Incineration (%) 66.1 70.3Waste to landfill (%) 74.5 72Landfill use (%) 128 94Failure in heavy duty equipment (nr./103 ton/year) 0.17 0.16Waste characterisation (�) 1.5 2.0Personnel (nr./103 ton/year) 0.35 0.58

Environmental sustainabilityLeachate tests performed (%) 90 98Leachate quality after treatment (%) 86 89Efficiency use of energy resources (kWh/ton) �66.5 �88.2Monitoring of groundwater quality (%) 84 90Monitoring of air emissions (%) 99.8 100

a Estimated.

provides some comments regarding each indicator with the aimof helping the operators to improve their quality of service. ERSARalso considers the adjustment for environment, by means of non-controllable factors that, somehow, may justify the (bad) perfor-mance determined for the indicator or the utility. Furthermore,the divergence between the result of the performance indicatorand the corresponding reference values is qualitatively classifiedas good, acceptable or unsatisfactory, through a green, yellow orred ball, respectively.

The sunshine regulation results have proven its benefits(Simões et al. 2010), which can be observed in Table 1. Duringthe presentation and public display of the results in 2004 (first yearof application) a great controversy arose in the sector, with theutilities questioning the methodology and pointing out the weak-nesses of the performance indicators as partial measures of pro-ductivity. If they were mostly right in their criticism, thejustification of the regulator was stronger since this discussionand controversy had instigated the improvement of the waste sec-tor performance. Nowadays, the sunshine regulatory model iswell-accepted by the industry and it is used by the media to pres-sure the firms.

4. Productivity analysis

4.1. Introduction

There are several methods to compute productivity change. Oneof the options relies on the assumption of a predefined production(or cost) technology function (parametric approach) that serves asbenchmark to measure the productivity scores but we need toknow the function specification. Other option is to apply non-parametric techniques, such as the Törnqvist Productivity Index(TPI) in quantities which, for example, measures the ratio of alloutputs (weighted by the corresponding revenues) to all inputs(weighted by cost shares) in quantities by using the observationsthemselves (waste utilities). The TPI compares the waste utilitywith its own performance, allowing for productivity measurementover time and is a non-frontier method. Non-parametric frontiermethods (e.g. the Malmquist Productivity Index – MPI or theHicks-Moorsteen Productivity Index – HMPI), although morecomplex to compute (using mathematical programming), do notrequire the inherent assumption of a function between the vari-ables, as the non-frontier methods do, nor a functional specifica-tion, like in the parametric methods (see Fried et al., 2008).

The MPI, being a frontier method, does not assume that all util-ities are cost minimisers or revenue maximisers. However, it hasother favourable features (Fried et al., 2008), namely, (a) the factof usually being based on non-parametric frontier models thatabolish the necessity of defining a production (or cost) function apriori, mitigating some inseparable errors; (b) the MPI calculationis based only on data on quantities; (c) it allows for the decompo-sition of the productivity in efficiency change (EffCh) and in varia-tions of technological change (TechCh).

Despite the wide application in the literature of the MPI inproductivity analysis, some issues remain to be solved, such asthe productivity interpretation, mainly under variable returns toscale (VRS), and infeasibilities (Epure et al., 2011). In this regard,Bjurek’s (1996) proposed the HMPI, which is defined as a ratio ofan aggregate output-quantity over an aggregate input-quantityindex. More precisely, it measures the change in output quantitiesin the output orientation and the change in input quantities inthe input orientation, instead of exclusively adopting an inputor output orientation as the MPI requires. Nevertheless, despiteits attractive properties, the HMPI has been scarcely appliedempirically.

1270 P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275

To determine the productivity of Portuguese waste utilities weapply the TPI, the MPI and the HMPI.

4.2. Törnqvist Productivity Index

The TPI is a weighted geometric average of the prices (p), withweights given by the simple average of the value shares at the peri-ods s (base) and t (current).

The Törnqvist quantity index (QT) formula is:

Q Tst ¼

YN

i¼1

qit

qis

� �WisþWit2

ð1Þ

where wis and wit present the value share of the i-commodity in thebase period (s), that is, pisqis=

PNi¼1pisqis:

4.3. Malmquist Productivity Index

The MPI is computed based on the efficiency measures achievedwith linear programming techniques, more precisely with distancefunctions (Di

t (y,x)) (see Shephard, 1953). Considering an input ori-entation (for output orientation the explanation is analogous), itcompares two vectors of inputs for a reference technology, byusing radial contraction of inputs, as follows:

MPIti ¼

Dti ðytþ1; xtþ1ÞDt

i ðyt ; xtÞð2Þ

Thus, the productivity change of a waste utility i is here mea-sured by the ratio of two distance functions, by using period t tech-nology as a reference. This productivity change is given by therelation between the distance function using period t + 1 relativeto the period t reference technology, [Dt

i ðytþ1; xtþ1Þ] and the dis-tance function using period t for the same period t reference tech-nology, [Dt

i ðyt ; xtÞ].MPI can be also written as (Färe et al., 1989):

MPIi ¼Dtþ1

i ðytþ1; xtþ1ÞDt

i ðyt ; xtÞ

!� Dt

i ðytþ1; xtþ1ÞDtþ1

i ðytþ1; xtþ1Þ

!� Dt

i ðyt; xtÞDtþ1

i ðyt ; xtÞ

!" #1=2

ð3Þ

The contribution of the EffCH (outside the square brackets) rep-resents the change of the production factors relative to the mini-mum inputs that still produce the desired outputs (efficientfrontier) in the time interval considered. The TechCh (within thesquare brackets of the equation) can be induced by an increase(or decrease) of the rate of transformation of inputs into outputs.Both indexes can take values greater, equal or lower than 1, repre-senting their improvement, maintenance or decline, respectively.

Färe et al. (1995) developed an approach to encompass thequality of service change in the productivity analysis. The authorsoffer a fairly general solution which can be applied to a wide set ofcontexts, including that of regulation. They present a MPI allowingfor the trade-offs between quality and quantity of operations,decomposing the TFP change into its quantity and quality compo-nents. Following the evolution over time of this index and itsdecomposition, the MPI illustrates the operators’ trade-off betweenquantity and quality outputs. In this regard, constant returns toscale (CRS) technology is also assumed.

The decomposition proposed by Färe et al. (1995) can be calcu-lated based on diverse distance functions combining informationof sequential periods, assuming a quantity index (QTtþ1

i ) and aquality index (QLtþ1

i ) , as follows:

MPIi ¼ QTtþ1i � QLtþ1

i ð4Þ

where,

QTtþ1i ¼ Dtþ1

i ðytþ1; qt; xtþ1ÞDtþ1

i ðyt ; qt; xtÞ

!� Dt

i ðytþ1; qtþ1; xtþ1ÞDt

i ðyt; qtþ1xtÞ

!" #1=2

ð5Þ

and,

QLtþ1i ¼ Dtþ1

i ðytþ1; qtþ1; xtþ1ÞDtþ1

i ðytþ1; qt; xtþ1Þ

!� Dt

i ðyt; qtþ1; xtÞDt

i ðyt ; qtxtÞ

!" #1=2

ð6Þ

4.4. Hicks-Moorsteen Productivity Index

In order to overcome some technology restrictions identified inthe MPI, we computed the HMPI. The HMPI is recognised as theonly multiplicatively-complete index that can be computed with-out price data, and, until now, it has not been applied to the wastesector (Epure et al., 2011). The HMPI is actually a ratio of(Malmquist) output and input quantity indexes.

Changing the MPI technology approach, which uses only inputor output orientation, the HMPI combines output and input quan-tity indices, using output and input distance functions respectively,becoming simultaneously input and output oriented (Bjurek,1996).

For a base period t, HMPI:

HMPITðtÞðxtþ1; ytþ1; xt ; ytÞ ¼Do

TðtÞðxt; ytÞ=DoTðtÞðxt ; ytþ1Þ

DiTðtÞðxt; ytÞ=Di

TðtÞðxtþ1; ytÞð7Þ

Assuming the same interpretation as for the MPI, when the out-put quantity index (ratio in numerator) is larger (smaller) than 1,then more (less) outputs are produced in period t + 1 than in periodt from a given input vector. When the input quantity index (ratio indenominator) is larger (smaller) than 1, then less (more) inputs areneeded in period t + 1 than in period t to produce a given outputvector. When the HMPI is larger (smaller) than 1, then it indicatesproductivity gains (losses).

In a similar way, a base period t + 1 HMPI index is defined asfollows:

HMPITðtþ1Þðxtþ1; ytþ1; xt ; ytÞ ¼Do

Tðtþ1Þðxtþ1; ytÞ=DoTðtþ1Þðxtþ1; ytþ1Þ

DiTðtþ1Þðxt ; ytþ1Þ=Di

Tðtþ1Þðxtþ1; ytþ1Þð8Þ

A geometric mean of these two HMPI indices yields:

HMPITðtÞ;Tðtþ1Þðxtþ1; ytþ1; xt ; ytÞ

¼ HMTðtÞðxtþ1; ytþ1; xt ; ytÞ � HMTðtþ1Þðxtþ1; ytþ1; xt; ytÞ� �1=2 ð9Þ

This one has the same interpretation as the previous index.

5. Productivity of Portuguese waste utilities

5.1. Data description and sample

In this paper, we focused on the ‘wholesale’ market since the‘retail’ market is still little regulated. The sample comprises anunbalanced panel data encompassing the 32 operating waste util-ities existent until 2010. This analysis includes the whole Portu-guese population. 17 waste utilities are regulated by ERSAR and15 are non-regulated by this sector specific regulator. Some ofthe latter (9) are subject to regulation by contract (those whichare public–private partnerships). The remaining utilities are onlysupervised by the municipalities.

In general, the data were obtained from each utility’s annual ac-count report from 2001 to 2008. However, in particular circum-stances, the data were checked and collected from ERSAR or bydirect contact with the waste utilities. The unbalanced data didnot affect the results since the absent waste utilities (always forthe first years) were considered to have a productivity equal to

P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275 1271

one (inputs and outputs equal to those of the first year with avail-able data) so that we could apply the frontier methods (MPI andHMPI).

5.2. Model specification

Regarding the model and variables adopted, there is no stan-dardisation for their selection. The option was always based onthe literature and on the authors’ expertise and experience on field,on the adequacy of the variables in economic terms and also on theavailable data. We opted to include three inputs and two outputs.The model uses the capital expenses (CAPEX), labour and the othercosts (OOPEX) as inputs. The outputs adopted are the waste treatedin the system (including non-hazardous industrial waste) and thewaste recycled, where different streams of waste (e.g. glass, paper,packaging, metal, greens, etc.) are considered. Note that being reg-ulated or non-regulated or having public or private ownership doesnot change the operating features of waste utilities and thereforethey are homogeneous and comparable with each other.

The index numbers require the specification of the variableweights, in opposition to the MPI which uses ‘‘optimal’’ or shadowweights for the data. The outputs are weighted through the reve-nues that each one of them produces: (a) the revenue from wastetreatment and (b) the revenue from waste recycled. The inputs areweighted through the corresponding costs. The labour is calculatedby the full-time employees (FTE). The cost of this input is the staffcost divided by the number of employees. CAPEX is measuredthrough the net assets and its price is established through the cap-ital cost, i.e. through the sum of the depreciation plus the provisionand interest expenses, conveyed as a percentage of the net assets.The last input refers to the OOPEX, which is measured by the totalcosts, subtracting the CAPEX and the staff costs. The weighted formadopted for this input is an implicit price deflator that reflects theconsumer price index (CPI) in the period studied. Both the capitalquantities and the OOPEX quantities are measured in monetaryunits with 2008 reference values, having been corrected for infla-tion (excluding housing).

Regarding the analysis of the quality of service improvement,we developed an index based on the sample of performance indi-cators used by ERSAR (Table 1). This indicator considers an equalweight for each performance indicator. Table 2 provides a snapshotof the statistical data used to compute the productivity indexes.

In the waste sector there are demand side management policiesand as the waste utilities must provide the service to all customers,the focus in terms of efficiency must be on optimising theproduction factors (inputs) rather than the results (outputs). So,

Table 2Average output and input quantity and price values for 2001 to 2008.

2001 (N = 12) 2008 (N = 32)CPI 0.844 1.000

Quantity Price Quantity Price

InputCAPEX

(103 €)/(-)5852.42 0.131 3174.98 0.179

Staff(-)/(€/emp.)

2137.33 22.70 1872.93 20.28

OOPEX(103 €)/(-)

5001.85 0.844 4696.68 1.000

OutputUrban solid waste treated

(ton)/(€/ton)287,155.82 22.07 179,050.94 35.51

Urban solid waste recycled(ton)/(€/ton)

9847.49 66.73 13,328.50 153.85

Quality Index 1.000 1.291

an input orientation was adopted in line with the literature andfollowing the principles of public service, therefore intending to re-duce costs rather than to increase profits (waste production in thiscase).

5.3. Results

Table 3 shows the results obtained from all three approaches. Atthe end of the period analysed, we observed cumulative values be-low 0.9 both by the TPI and the MPI. This means that after 8 yearsthe waste utilities ended up with a productivity decline of morethan 10%. Both techniques led to this conclusion. Although thehighest productivity progress was observed in 2005 (TPI = 1.056,MPI = 1.023, HMPI = 1.022), this was not enough to contradict thegeneral trend of productivity decline. It was also noticed that theyear of 2008 was the major contributor to the entities productivitydecline (TPI = 0.936, MPI = 0.952, HMPI = 0.951). 2005 was an atyp-ical year for two reasons. First, it was the first year that the resultsof regulation were felt and therefore they benefited the productiv-ity of regulated waste utilities. Second, that year was an electionyear and, therefore, the expenses of non-regulated municipal com-panies were above the normal and penalised their productivity.

During the period under analysis, only in 2005 was there pro-ductivity progress among the waste entities. The other years werecharacterised by a constant productivity decline. In Fig. 1 it is pos-sible to observe the evolution of the TPI, MPI and HMPI averages, aswell as minimum and maximum values.

Table 4 presents the productivity (accumulated) results of reg-ulated (Reg) and non-regulated (NReg) utilities.

Fig. 2 shows the efficiency and technological change obtainedfrom the MPI computation.

From Table 3 and Figs. 1 and 2, we can observe that, on average,the waste utilities show a productivity decrease of more than 1%per year. Considering the institutional environment change, withmore and more demanding legislation (including the sunshine reg-ulatory model that was implemented), the major contributor toproductivity change was the technological change, as expected.Despite some improvement in efficiency, it did not compensatethe severe decline in technological change. This means that thePortuguese waste utilities have been moving their production lev-els closer to the best-practice production frontier. However, thefrontier itself has shifted farther away, entailing technology changedecline. This is in line with the expected since the waste sector has,by nature, a negative technological change due to strict legislationand demanding requirements imposed over time.

6. Hypotheses analysis

As mentioned, our aim was to evaluate the influence of the reg-ulatory model on the performance of the urban waste utilitiesactivity, in particular, the results and incentives provided by sun-shine regulation concerning the quality of service, and of RoRand price cap for economic regulation. Different research questionswere raised related to the influence of these aspects. They are pre-sented next together with the answers:

6.1. Has the ‘sunshine regulation’ method produced positive results inthe quality of service regulation?

It is widely recognised that the ‘name and shaming’ strategiespromoted by sunshine regulation have been very effective as faras quality of service regulation is concerned (Marques and Simões,2008). The results (see also Table 1) provide evidence of such effec-tiveness, as can be observed by Fig. 3 when an aggregated index(considering equivalent weights) was used.

Table 3Productivity cumulative indexes at the period 2002–2008.

TPI MPI HM

Average values Cumulative values Average values Cumulative values Average values Cumulative values

2002 0.995 0.995 0.989 0.996 0.988 0.9882003 0.944 0.941 0.959 0.955 0.965 0.9602004 0.987 0.929 0.962 0.925 0.965 0.9322005 1.056 0.979 1.023 0.949 1.022 0.9572006 1.008 0.965 0.973 0.899 0.973 0.9062007 0.994 0.959 0.978 0.896 0.980 0.9002008 0.936 0.894 0.952 0.863 0.951 0.864Average 0.989 0.977 0.978

Fig. 1. Average, maximum and minimum values for TPI, MPI and HM.

Table 4Productivity results (accumulated) for regulated and non-regulated utilities (from2002 to 2008.

TPI MPI HMPI

Reg Nreg Reg Nreg Reg Nreg

2002 1.028 0.986 0.995 0.968 0.995 0.9732008 0.887 0.902 0.873 0.884 0.869 0.887

1272 P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275

The results proved the benefits of sunshine regulation in termsof quality of service improvement. The quality index (QL) of theMPI depicts an improvement of about 10% of the utilities regulated.

Moreover, it is interesting to realise that the highest slope inthis analysis is observed right after the implementation of ERSAR’ssunshine regulation (in 2004). The quality of service improvementin regulated utilities is even more relevant when compared withthe non-regulated ones, where only a slight improvement of qual-ity of service is noticed (Fig. 3). This is a clear evidence of the re-

Fig. 2. Decomposition of the MPI.

Fig. 3. Weighted index of quality of service for regulated and non regulatedutilities.

Fig. 4. Weighted index of quantity for regulated and non-regulated utilities.

P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275 1273

sults and incentives provided by sunshine regulation to the wasteindustry.

6.2. Does the ‘sunshine regulation’ method have a perverse effect onthe utility productivity?

The role of sunshine regulation regarding the quality of serviceimprovement is well documented. Nevertheless, it is ‘‘not easy tohave the best of both worlds’’, meaning that the results about thequality of service require extra investments and expenses fromthe utilities, which surely influence their total costs, and thereforetheir productivity. Fig. 4 shows the productivity evolution for theutilities under ‘‘sunshine regulation’’ and for non-regulated utili-ties. The results are clear, however, in 2005, there was a kind of

Fig. 5. MPI decomposition results for the

inversion of the results which can be related to the elections atthe municipal level that triggered some lack of financial controlby non-regulated municipal companies. At the same time, somecosts were rejected by the regulator in several waste utilities dur-ing the RoR process.

It is also recognised that, in terms of economic regulation, sun-shine regulation has limited powers. As expected, the regulatedutilities show a productivity decline, mainly due to the externalcost required in investments and other expenses to counterbalancethe incentives induced in terms of quality of service. The non-reg-ulated utilities also show a drop in productivity, but to a lessextent.

6.3. Is the low productivity change compensated by the improvementin the quality of service?

The resources needed to improve the quality of service raisedsome questions about their implications on the utilities productiv-ity. In this regard, we investigated the trade-off between utilitiesproductivity (QT) and quality of service (QL) through the decompo-sition of the MPI (Färe et al., 1995). Fig. 5 shows the results ob-tained for the regulated and non-regulated utilities.

From Fig. 5, the ‘‘non-compensation’’ of the quality of service in-crease in terms of productivity analysis is evident. We clearly ob-serve that the improvement of the quality index does not justifythe decline of the quantity index. This result cannot be comparedwith that of the non-regulated utilities (Fig. 5).

Despite the good results pointed out at the level of the quality ofservice enhancement in Portugal with sunshine regulation, untilnow the benefits of productivity improvement have not been clear.Indeed, these results are not being accompanied by the same kindof success in economic terms. Moreover, this situation raises some

regulated and non-regulated utilities.

Fig. 6. Productivity evolution of utilities under RoR and price cap. Fig. 7. MPI results of utilities under RoR.

1 We should note that ‘gold plating’ practices refer basically to investments that areunnecessary and in the Portuguese case-study some of them in fact corresponded tothe real improvements. However, we believe the amount of investments wasexcessive without provide value for money. Although our analysis is fundamentallyeconomic and only one of the productivity methods used controlsthe quality, wedefend that the appropriate quality of service should be imposed and that too muchquality of service might not be good since we must pay for it.

1274 P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275

questions about the customers’ willingness to pay for this level ofquality of service.

6.4. Does RoR have a negative effect in the productivity change,compared with the other economic regulatory methods?

RoR is characterised by a short regulatory period and a low riskprofile, but encourages the steadiness and ‘quiet-life’ of managersand X-inefficiency. Additionally, it fosters overinvestment andgold-plating practices. This circumstance definitely raises somequestions about the (economic) incentives that are being sent tothe market. Fig. 6 shows the effect of RoR (for the regulated utili-ties) and price cap (for the non-regulated utilities subject to regu-lation by contract) on the productivity change over time.

Fig. 6 highlights the incentives behind these two economic reg-ulatory approaches. Regulation by contract based on maximumprices yields better results than RoR in terms of productivity. Infact, after the eight-year period of analysis, the utilities underRoR had lost more than 5% compared with the ones under pricecap. Indeed, under RoR the utilities are not encouraged to be effi-cient and innovative. The least evident effect noticed in 2005, asabovementioned, was related to the rejection of some costs bythe regulator during the RoR process.

In addition, Fig. 6 provides evidence of two other important as-pects also recognised in the literature review (Simões andMarques, 2011) which are: (a) the initial benefits of privatisationand their decline over time, and (b) the incentives provided bythe implementation of a sector-specific regulator.

6.5. Is the mixture of RoR and sunshine regulation risky in terms of thescarce incentives provided and value for money provided?

Extra investment and additional costs might be stimulated bysunshine regulation and RoR. This circumstance may be penalisingfor the productivity of waste utilities, with consequences for cus-tomers or taxpayers. Moreover, some questions are once againraised about the incentives provided by RoR to the market. Fig. 7shows the decomposition of MPI in quality of service (QL) and util-ities productivity (QT) only for the utilities subject to RoR.

Fig. 7 stresses the quality of service improvement, as a result ofthe ‘‘name and shaming’’ incentives of sunshine regulation.However, this outcome does not compensate the trade-off gener-ated between the extra-costs required to fulfil quality of servicerequirements and the (reduced) incentives promoted by RoR.

This circumstance suggests the idea that sunshine regulationshould be complemented by an economic regulation model basedon incentives (e.g. price cap or revenue cap regulation). For in-stance, in the water sector there are diverse studies proving its

importance (Aubert and Reynaud, 2005). RoR together with sun-shine regulation encourage overinvestment and additional costs,especially in infrastructure sectors where huge investments are re-quired and which are popular for politicians. Furthermore, in thesesectors the operators usually have close relationships withcontractors who are the main stakeholders interested in overin-vestment and gold-plating practices.1

7. Concluding remarks

The present study intends to evaluate the influence of regula-tion on the performance of the urban waste utilities. For this pur-pose, the productivity change of Portuguese waste utilitiesbetween 2001 and 2008 was measured by three different tech-niques. Particularly, we decomposed the MPI in quality and quan-tity indexes.

We concluded that the sunshine regulatory model is very posi-tive as far as quality of service improvement is concerned. We alsoconcluded that there was a productivity drop in that period, whichcan be related to overinvestment and the increase of expensesneeded for the improvement of the quality of service. This argu-ment was demonstrated when the MPI was decomposed, sinceboth the productivity decline and the improvement of the qualityof service are more visible in the regulated utilities than in thenon-regulated ones. We also compared the economic regulatorymethods. It seems evident that regulated utilities subject to RoRhave a higher drop in the productivity when compared with thenon-regulated utilities (subject to a maximum price in the regula-tory contracts).

Taking into account the effect of sunshine regulation and RoRtogether, we can infer that practices of gold-plating and overin-vestment might be implemented in the Portuguese waste sectorand customers and/or taxpayers might be penalised if they arenot willing to pay for the higher quality attained. Therefore, wethink that sunshine regulation should be promoted and imple-mented by regulators but as a complement to other (economic)performance based regulatory method that fosters incentives forefficiency and innovation among waste utilities. This conclusionis demonstrated by the empirical evidence of regulators of waterutilities in some States in Australia (e.g. ESC, previous ORG) which,after a certain period applying only sunshine regulation, comple-

P. Simões, R.C. Marques / Waste Management 32 (2012) 1266–1275 1275

mented it with the implementation of economic regulation for thereasons here pointed out.

References

Ai, C., Sappington, D., 2002. The impact of state incentive regulation on the U.S.telecommunications industry. Journal of Regulatory Economics 22 (2), 133–160.

Aubert, C., Reynaud, A., 2005. The impact of regulation on cost efficiency: anempirical analysis of Wisconsin water utilities. Journal of Productivity Analysis23 (3), 383–409.

Bel, G., Costas, A., 2006. Do public sector reforms get rusty? Local privatization inSpain. Journal of Policy Reform 9 (1), 1–24.

Benfratello, L., 2009. Technology and incentive regulation in the Italian motorwaysindustry. Journal of Regulatory Economics 35 (2), 201–221.

Benito, B., Bastida, F., García, J., 2010. The determinants of efficiency in municipalgovernments. Applied Economics 42 (4), 515–528.

Berg, S., Lin, C., Tsaplin, V., 2005. Regulation of state-owned and privatized utilities:Ukraine electricity distribution company performance. Journal of RegulatoryEconomics 28 (3), 259–287.

Bjurek, H., 1996. The Malmquist total factor productivity index. ScandinavianJournal of Economics 98 (2), 303–313.

Braconnier, S., 2001. La regulation des services publiques. Revue française de droitadministrative 17 (1), 43–54.

Bosch, N., Pedraja, F., Suarez-Pandiello, J., 2000. Measuring the efficiency of Spanishmunicipal refuse collection services. Local Government Studies 26 (3), 71–90.

Cambini, C., Rondi, L., 2009. Incentive regulation and investment: evidence fromEuropean energy utilities. Journal of Regulatory Economics 38 (1), 1–26.

Chapple, W., Harris, R., 2003. Accounting for solid waste generation in measures ofregional productivity. Research Paper Series of ICCSR. Reports.

De Jaeger, S., Eyckmans, J., Rogge, N., Van Puyenbroeck, T., 2011. Wasteful waste-reducing policies? The impact of waste reduction policy instruments oncollection and processing costs of municipal solid waste. Waste Management31 (7), 1429–1440.

Dijkgraaf, E., Gradus, R., 2003. Cost-savings of contracting out refuse collection.Empirica 30 (2), 149–161.

Domberger, S., Meadowcroft, S.A., Thompson, D.J., 1986. Competitive tendering andefficiency: the case of refuse collection. Fiscal Studies 7 (4), 32–44.

Eckenrod, S., 1999. Incentive regulation in local telecommunications: the effects onprice markups. Journal of Regulatory Economics 30 (2), 217–231.

Epure, M., Kerstens, K., Prior, D., 2011. Technology-based total factor productivityand benchmarking: new proposals and an application. OMEGA – TheInternational Journal of Management Science (forthcoming).

Estache, A., Perelman, S., Trujillo, L., 2007. Measuring quantity-quality trade-offs inregulation: the Brazilian freight railways case. Annals of Public and CooperativeEconomics 78 (1), 1–20.

Estache, A., Rossi, M., 2005. Do regulation and ownership drive the efficiency ofelectricity distribution? Evidence from Latin America Economics Letters 86 (2),253–257.

Färe, R., Grosskopf, S., Roos, P., 1995. Productivity and quality changes in Swedishpharmacies. International Journal of Production Economics 39 (1/2), 137–144.

Färe, R., Grosskopf, S., Lindgren, B., Roos, P., 1989. Productivity developments inSwedish hospitals: a Malmquist output index approach. In: Discussion PaperSeries, 89, Carbondale: Southern Illinois University at Carbondale.

Fleishman, R., Alexander, R., Bretshneider, S., Popp, D., 2009. Does regulationstimulate productivity? Effect of air quality policies on efficiency of US powerplants. Energy Policy 37 (11), 4574–4582.

Fried, H., Lovell, C., Schmidt, S., 2008. The Measurement of Productive Efficiency andProductivity Change. Oxford University Press, New York.

Giannakis, D., Jamasb, T., Pollitt, M., 2005. Benchmarking and incentive regulation ofquality of service. an application to UK electricity distribution network. EnergyPolicy 33 (17), 2256–2271.

Gómez-Lobo, A., Szymanski, S., 2001. A law of large numbers: bidding andcompulsory competitive tendering for refuse collection contracts. Review ofIndustrial Organization 18 (1), 105–113.

Hammond, C., Jones, G., Robinson, T., 2002. Technical efficiency under alternativeregulatory regimes: evidence from the inter-war British gas industry. Journal ofRegulatory Economics 22 (3), 251–270.

Ida, T., Ito, E., Kinoshita, S., 2007. Post-regulatory reform productivity gains inJapan’s electricity industry. Applied Economic Letters 14 (13), 975–979.

Jamison, M., Berg, S., 2004. Annotated reading list for the body of knowledge oninfrastructure regulation. Interdisciplinary Studies in Literature andEnvironment 11 (1), 281–292.

Kahn, A., 1988. The Economics of Regulation: Principles and Institutions. MIT Press,Cambridge, MA.

Lam, P., Lam, T., 2005. Total factor productivity measures for Hong Kong telephone.Telecommunications Policy 29 (1), 53–68.

Lin, C., Berg, S., 2008. Incorporating service quality into yardstick regulation: anapplication to the Peru water sector. Review of Industrial Organization 32 (1),53–75.

Margari, B., Erbetta, F., Petraglia, C., Piacenza, M., 2007. Regulatory andenvironmental effects on public transit efficiency: a mixed DEA-SFA approach.Journal of Regulatory Economics 32 (2), 131–151.

Marques, R., 2006. A yardstick competition model for Portuguese water andsewerage services regulation. Utilities Policy 14 (3), 175–184.

Marques, R., Simões, P., 2008. Does the sunshine regulatory approach work?Governance and regulation model of the urban waste services in Portugal.Resources, Conservation & Recycling 52 (8/9), 1040–1049.

Maziotis, A., Saal, D., Thanassoulis, E., 2009. Regulatory price performance, excesscost indexes and profitability: how effective is price cap regulation in the waterindustry? Aston University Working Paper 920, 42 p.

Minken, H., Killi, M., 2000. Productivity growth in Norwegian ferry links 1988–1996, and applications for regulation. TØI-rapport 482, Oslo.

Mizutani, F., Kozumi, H., Matsushima, N., 2009. Does yardstick regulation reallywork? Empirical evidence from Japan’s rail industry. Journal of RegulatoryEconomics 9 (3), 269–306.

Motta, R., Moreira, A., 2006. Efficiency and regulation in the sanitation sector inBrazil. Utilities Policy 14 (3), 185–195.

Nakano, M., Managi, S., 2008. Regulatory reforms and productivity: an empiricalanalysis of the Japanese electricity industry. Energy Policy 36 (1), 201–209.

Odeck, J., 2008. The effect of mergers on efficiency and productivity of publictransport services. Transportation Research Part A 42 (4), 696–708.

Ohlsson, H., 2003. Ownership and production costs. Choosing between publicproduction and contracting-out in the case of Swedish refuse collection. FiscalStudies 24 (4), 451–476.

Orr, F., Lefebvre, B., 1993. The impact of regulation on telecommunicationsproductivity and price performance. Utilities Policy 3 (4), 311–320.

Pássaro, D., 2003. Report: waste management in Portugal between 1996 and 2002.Waste management 23 (1), 97–99.

Pérez-Reyes, R., Tovar, B., 2009. Measuring efficiency and productivity change (PTF)in the Peruvian electricity distribution companies after reforms. Energy Policy37 (6), 2249–2261.

Resende, M., Facanha, L., 2005. Price-cap regulation and service-quality intelecommunications: an empirical study. Information Economics and Policy17 (1), 1–12.

Roycroft, T., 1999. Alternative regulation and the efficiency of local exchangecarriers: evidence from the Ameritech states. Telecommunications Policy 23 (6),469–480.

Saal, D., Parker, D., 2001. Productivity and price performance in the privatized waterand sewerage companies of England and Wales. Journal of RegulatoryEconomics 20 (1), 61–90.

Sanchez, I., 2008. The performance of solid waste collection in Spain. WasteManagement & Research 26 (3), 327–336.

Santos, J., Furtado, A., Marques, R., 2010. Reform and regulation of the Portugueserail sector. What has failed? Utilities Policy 18 (2), 94–102.

Sanyal, P., 2007. The effect of deregulation on environmental research by electricutilities. Journal of RegulatoryEconomics 31 (3), 335–353.

Seo, D., Shin, J., 2011. The impact of incentive regulation on productivity in the UStelecommunications industry: a stochastic frontier approach. InformationEconomics and Policy 23 (1), 3–11.

Shephard, R., 1953. Cost and Production Functions. Princeton University Press,Princeton.

Simões, P., Marques, R., 2011. On the cost inefficiency of the waste sector. Aliterature review. Working Paper, IST, Lisbon.

Simões, P., De Witte, K., Marques, R., 2010. Regulatory structures and operationalenvironment in the Portuguese waste sector. Waste Management 30, 1130–1137.

Spasovic, L., Sideris, A., Das, S., Chao, X., 1999. Increasing productivity and servicequality of the straddle carrier operations at a container port terminal. Journal ofAdvanced Transportation 20, 1999.

Ter-Martirosyan, A., 2003.The effects of incentive regulation on quality of service inelectricity markets. Working Paper. George Washington University, USA.

Uri, N., 2002. Assessing the effect of incentive regulation on productivity efficiencyin telecommunications in the United States. European Journal of Law andEconomics 13 (2), 113–127.

Wilson, W., 1997. Cost savings and productivity in the railroad industry. Journal ofRegulatory Economics 11 (1), 21–40.

Witte, K., Saal, D., 2010. Is a little sunshine all we need? On the impact of sunshineregulation on profits, productivity and prices in the Dutch drinking watersector. Journal of Regulatory Economics 37 (3), 219–242.

Worthington, A., Dollery, B., 2001. Measuring efficiency in local government: ananalysis of New South Wales municipalities’ domestic waste managementfunction. Policy Studies Journal 29 (2), 232–249.

Zhang, Y., Parker, D., Kirkpatrick, C., 2008. Electricity sector reform in developingcountries: an econometric assessment of the effects of privatization,competition and regulation. Journal of Regulatory Economics 33 (2), 159–178.