Does the political and economic context influence the success of a transport project? an analysis of transport public-private partnerships

Does the political and economic context influence the success of a transportproject? An analysis of transport public-private partnerships a Department of Transport Engineering and Logistics, Pontificia Universidad Católica de Chile, Casilla 306, Cod. 105, Santiago 22, Chile b UCL QASER Lab, University College London, Gower Street, London WC1E 6BT, UK The construction and provision of infrastructure services such as transport nowadays is often based on a partnership between three main actors: public sector, private sector and multilateral lenders, under a framework of PublicePrivate Partnerships (PPPs). This type of partnership has been employed in a wide range of projects in the transport sector and in various contexts in developing and developed countries.
Given this observation, the objective of this paper is to examine how countries’ economic and politicalcharacteristics contribute to the success of PPPs in transport investments. Special focus in the analysis isgiven to how the perception of corruption and democratic accountability may influence the success ofa PPP project in different transport sectors. We examine a database with 856 transport PPP projects usinga generalized linear model in the form of a logit model in order to evaluate the transport databasecovering data from 72 countries, classified in six regions. The study highlights the importance of nationalexperience. Not only does national macroeconomic experience appear to have a relevant role, but so alsodoes its past experience (either positive or negative) of transport PPP projects. An interesting findingfrom the analysis is the importance of the rest of the world’s perception of a country’s level of corruptionand democratic accountability for the final outcome of a PPP project.
can be replicated across transport sectors and across countries. Thechoice context is indeed a multi-objective decision, and in practice, National governments around the world differ substantially in the three actors have to reach a judgment about the trade-offs their social and economic structures and in particular in their between the various, sometimes conflicting, objectives.
infrastructure endowments. State governments are characterized The literature devotes special attention to the difficulties in PPP by diverse administrative cultures and capabilities and distinct legal agreements between the public and private sector ( and planning traditions. For instance, institutional diversity in the transport sector is considerable, with countries adopting different banks are seen as the party that always wins even if approaches with respect to user charges and ownership structures, a project fails, or if the government and the private company have and thereby implementing various approaches to infrastructure to renegotiate the PPP. Within this framework, multilateral lenders investment strategy and financing. Despite these differences, such as the European Union and the World Bank have openly a framework for what are now referred to as PPPs (PrivateePublic supported public projects involving PPP agreements between Partnerships) has emerged to provide transport services through private investors and governments, especially from developing partnerships between three main actors: public sector, private sector and multilateral lenders. The main potential benefit of the number of papers analyze the behaviour of the private investor, in PPP approach in transport is its flexibility in adapting the structure particular by focusing on the maximization of private benefit under of incentives and risk-sharing to the features of the project and to the economic and institutional environment. But because of this flexibility, it is perhaps unwise to seek a unique model of PPP that When examining PPP agreements, several authors observe the necessity for a shift in the public sector role: that is, from beingmerely a provider to increasingly becoming a regulator * Corresponding author. Tel.: þ56 (2) 354 4270.
E-mail addresses: (P. Galilea), (F. Medda).
a legislative and administrative framework in order to facilitate PPP 0739-8859/$ e see front matter Ó 2010 Published by Elsevier Ltd.
P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 related to transport. It reflects not only the government’s reputation use PPP arrangements, we observe different ways of adopting this in its capacity to honour agreements with the private sector, but approach due to different cultural influences and traditions in plan- also the capability of the private sector to accomplish projects with ning and management of public works, deficiencies in legal and the private sector. This experience has proven to be a critical institutional structures, and different degrees of political awareness predictor of successful future PPP arrangements ( ). Positive outcomes and thus country experiences on highlight the potential significance of a country’s previous transport PPPs are associated with positive outcomes of past experience in PPPs in attracting further PPP projects to that country. However, we observe that there is as yet no empirical Hypothesis 1a. Successful country experience on previous evidence showing how this experience may (or may not) affect later transport PPP projects is positively associated with the outcome of PPP outcomes. Also, the connection between a country’s level of corruption has not been studied in the light of its influence in thesuccess of a transport PPP project. Several studies have been made Past experience sometimes also implies the existence of about corruption and its influence on economic growth unsuccessful PPP projects. This experience, although “bad”, might enhance the future chances for successful PPP projects due to lessons learned from a negative experience. However, we assume but none has been conducted using a stringent microeconomic here that having unsuccessful PPP projects means having a black spot on a country’s record of PPP projects, and can therefore The objective of the present paper is to examine how the three potentially discourage future private investments, attract fewer actors, public sector, private sector and multilateral lenders, each investors, and may also signal to the government or the public contributes to the success of PPPs in transport investments, by sector that they are not coping successfully with PPP projects.
considering different political and socioeconomic contexts. We will Hypothesis 1b. Unsuccessful country experience on previous also focus our analysis on the effect of a country’s level of corrup- transport PPP projects is negatively associated with the outcome of tion and democratic accountability in the success of a PPP project.
The paper is organized as follows: Section presents our hypotheses with their theoretical backgrounds. In Section , wedescribe the dataset used to test the hypotheses previously 2.1.2. Country’s macroeconomic performance described, outline the dependent and independent variables The stability of a country, based on its macroeconomic condi- employed in our analysis, and explain the modelling procedure.
tions, is important in order to attract private and foreign investors Section describes and analyzes our results on the variables that (especially in emerging markets, as shown in ), may affect a PPP outcome and thus concludes the paper.
and has also proved to be important in limiting the number of PPPsin a country (). We will analyze its effects on the positive outcome of a PPP. Poor macroeconomic conditions mayhinder the success of a PPP project, whereas a good macroeconomic In order to address the impact of the three actors on the success performance may foster better outcomes.
of PPPs, in this section we discuss the hypotheses that represent the Hypothesis 1c. Satisfactory country macroeconomic conditions backbone of our analysis. Although there are many elements which are positively related with the chances of successful PPP projects in influence the success of PPP agreements, we consider in this analysis three main building blocks: country experience, investorsand multilateral lenders.
The first block represents the country’s past experience in transport PPP projects as well as its macroeconomic performance Most of the economic literature agrees that corruption would when the project started and the way a country is perceived in terms of corruption and democratic accountability. This block will be the foundation for the success of the project and will (or will not) reinforce the subsequent blocks. We assume that a country with reduce economic growth as it lowers the incentive for entrepre- “bad” past experience in PPP projects and/or deficient macroeco- neurs to invest. Corruption can also distort the composition of nomic performance will not attract as many private investors for its government expenditure, shifting the expenditure of public PPP projects, as would another country with better experience. The resources from socially desirable projects to projects where it is second block is the link between the private investors involved and easier to extract large bribes. When a country is perceived as the PPP project. The private investor might have several charac- corrupt, there might be fewer private investors willing to support teristics, and in this paper we focus on the number of private projects in that particular country, constraining the set of potential investors forming the consortium in charge of the PPP project. The investors (and thus restraining the “optimal” investor for the final block represents the multilateral lenders supporting the PPP project). There is also a higher probability that the chosen provider project. Although some of the literature discusses their role as may not be the most capable, but rather the one with the best bribe, agents of policy change and focuses on how they might add thus limiting the likelihood for a successful outcome.
a degree of external coercive pressure to the PPP project’s national Hypothesis 1d. The more a country is perceived as corrupted, the less likely it is that the PPP has a positive outcome.
their presence as a means of success for the PPP project.
2.1.1. Country’s past experience with transport PPPs Past experience in running infrastructure projects related to 2 Some authors have pointed out that some level of corruption is desirable transport projects may be a good forecaster of future PPP outcomes P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 In order to test if the perception of corruption may be more Hypothesis 1i. The region where the project is located, and the relevant in some regions rather than in others, the interaction type of project affect the outcome of a PPP project.
Hypothesis 1e. The effect of the perception of corruption on the success of a PPP varies within projects in different regions.
As the number of private investors increases, it may be harder 2.1.4. Country’s democratic accountability index for them to agree and to work efficiently; therefore, a negative When a developing country is perceived as having low demo- outcome for the PPP project may become increasingly likely with cratic accountability (DA), it means that that country’s government is less responsive to its people. For instance, an autarchy would beperceived as having the lowest DA, whereas an alternating Hypothesis 2a. If there is more than one private investor on a PPP democrwould be perceived with the highest. Although it project, it is more likely that the PPP has a negative outcome.
might be the case that a lower number of investors would like to However, the ways the number of investors affect countries with invest in a country with a low DA, once a willing private investor is different incomes can differ. In countries with low- and lower selected for a PPP project, government support (with all its middle-incomes, more than one private investor in a PPP project authority) will follow, and so it is less likely that this PPP will fail.
could indicate that the consortium has broader expertise and Conversely, a PPP agreement in a country with a high DA will have proficiency in PPP projects; they will share part of the costs and government support, but it might be subjected to a shift in support risks; and more parties will be watchful of their own (and their due to change unforeseen by means of a democratic vote.
partners’) investments. These characteristics may prove to be more Hypothesis 1f. The more a country is perceived as having low relevant in countries with low- and lower middle-incomes, since democratic accountability, the more it is likely that the PPP has the country itself might not have the expertise on infrastructure investments. A positive outcome for the PPP project may thereforebe more likely with more than one private investor in low- and The influence of the perception of DA may differ among the different types of projects. Projects such as airports, seaports andrailroads are more capital-intensive than toll roads, thus they have Hypothesis 2b. If there is more than one private investor on a PPP a higher level of risk. Governments with lower DA will have more project, it is more likely that the PPP has a positive outcome in authority to assist these types of projects if needed, whereas a low- or lower middle-income country.
governments with higher DA will generally not be able to do it.
2.2.2. Private percentage of the project contract or company owned Hypothesis 1g. The effect of the perception of democratic accountability varies within different types of projects, thus Ownership is a major factor in the PPP literature, as discussed by affecting the final outcome of a PPP project.
and , because ownership willprovide certain incentives to the private sector in charge of the PPP project. In every PPP project a company is in charge of its devel- Countries belonging to certain regions usually share cultural, opment, or a project contract stipulates the rights and duties of the socioeconomic and political characteristics. They might have private parties. A project company may be owned by a percentage a similar rule of law, or they might react the same way to certain of the private investWhenever private investors own a larger situations or problems. There are also regions with more expe- share of the project company, they should have a greater incentive rience in PPP projects than others, as shown in to become involved and closely follow the results of the project.
Thus when they own a greater share, it is expected that a better region have received 50 percent (US$345 billion) of worldwide outcome can be achieved by the PPP project.
private capital flows to the infrastructure sectors during the Hypothesis 2c. Positive PPP outcomes are more likely to occur 1990s. The implication here is that the region where the project is when the private percentage ownership of the project company (or located can possibly affect the success of a PPP transport project.
Hypothesis 1h. The region where the project is located affects theoutcome of a PPP project.
Different types of projects may have diverse results among the Multilateral lenders or lending agencies (the World Bank, profitability of infrastructure concessions in Latin America and European Investment Bank and Asian Development Bank, among found differences among sectors. The experience that a region others) are sometimes involved in PPP projects by executing their has in toll roads versus seaports can be dissimilar, and the ways role as the giver of loans. As proven by the different societies might welcome certain projects can vary. The lending by these agencies stimulates growth in the recip- interaction between types of projects and interaction among the ient countries in some cases. To be sponsored by these multilateral lenders, government and private investors in a PPP project mustfulfil several conditions, such as the timing of recouping theinvestment, interest rates and regulation regime. Many lendersmonitor the PPP process from its inception, through the selection of 3 By alternative democracy we refer to a country’s democracy, where besides having fair and free elections to the executive and legislative powers, and an activepresence of more than one political party, there is a viable opposition and theexecutive power has not served more than two successive terms. In other words, it 4 The private investor’s ownership may be a percentage of the project contract or is a democracy where the same party or coalition has not been continuously in project company but this does not necessarily indicate the ownership of the P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 Status of the projects in the database.
Source: private participation in infrastructure projects database, World Bank.
The database only includes projects awarded in low- and Financial closure year
middle-income countries as classified by the World Bank (2005).
Source: Private Participation in Infrastructure Projects Database, World Bank The transport database covers data from 72 countries, classified insix regions: East Asia and the Pacific, Europe and Central Asia, Latin Fig. 1. Number of transport projects by year.
America and the Caribbean, the Middle East and North Africa, SouthAsia, and Sub-Saharan Africa. Almost half of the projects (44%) arefrom Latin America and the Caribbean, and dispersed mostly the private investor, and to its final development and completion. If among Brazil, Mexico, Argentina, and Chile. The projects from East a PPP project is sponsored by a multilateral lender, it will be Asia and the Pacific are highly concentrated in China, while the invigilated, thus the PPP’s failure should be increasingly unlikely.
projects from South Asia are concentrated in India.
Hypothesis 3. Existence of multilateral lenders in a PPP projectwill enhance the positive outcome of that PPP.
In the next section we will describe the dataset used to test the six hypotheses described, and we will explain the modelling Our dependent variable (Success) is a binary variable, taking the procedure and the dependent and independent variables employed value zero, if the project’s status was either cancelled or distressed, and one, if the project’s status was under construction, operationalor concluded. In order to estimate the regressions, we usea generalized linear model in the form of a logit model () To test the previous hypotheses, a database with 856 transport where Y is the dependent variable, x is the vector of independent PPP projects was used. The database is part of the Private Partici- variables, and b is the vector of parameters.
pation in Infrastructure Projects Database, which has projects fromfour sectors: energy, telecommunications, transport, and water. The original database is a joint product between the World Bank andthe Public-Private Infrastructure Advisory Facility In order Each project of the database may be in one of the following five to be included in the database, the project must involve the states: i) under construction (projects for which assets are being ownership or operation of physical assets required to provide the built); ii) operational (projects that have begun providing services infrastructure services, and must have a private investor who bears to the public); iii) concluded (projects for which the contract period a share of the project’s operational risk. Only 856 projects related has expired and the project was neither renewed nor extended by with the transport sector are analyzed in this paper. Transport either the government or the operator); iv) cancelled (projects from sector projects are divided into four subsectors: toll roads (47%), which the private sector has exited before the end stipulated in the seaports (29%), airports (13%), and railroads (11%).
contract); and v) distressed (projects where the government or the The databaseprovides information for transport projects that operator has either requested contract termination or are in have reached their financial closurebetween 1984 and 2005. illustrates that almost one-third of the projects reported reached The status of the project was grouped into a dichotomous their financial closure between 1996 and 1998. The increase in the measure, entitled Success, equal to one if the project’s status was number of projects reflected in 1990 is due mainly to the toll roads under construction, operational or concluded. In our sample of 856 subsector, whereas the increase until 1998, and the decline since projects, 804 were in this status (94%). If the project’s status was 1999, is reflected in all subsectors.
either cancelled or distressed, the dependent variable was set equalto zero. illustrates the total status of the projects in thedatabase and their relation to the dependent variable.
5 The database can be downloaded at .
6 Data regarding macroeconomic information for the countries included in the database was collected from the World Economic Outlook of the International Data regarding the corruption and democratic accountability index are Two variables measuring the past experience of a country in Financial closure, as defined by the Private Participation in Infrastructure transport PPPs were created, entitled Yes PPP Experience and No PPP Database, occurs when there is a legally binding commitment of private investors tomobilize funding or provide services.
Experience, respectively. For a PPP, Yes PPP Experience counts the P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 number of transport PPP projects done in the country of the PPP at the moment of the PPP’s financial closure; whereas No Incidence of multiple private investors across regions.
PPP Experience counts the number of unsuccessfultransport PPP projects done in the country of the PPP at the time of the PPP’s financial closure. Both variables are set to zero for countries with no prior experience in transport PPPs. PPP projects undertaken in the same country do not necessarily have the same values in Yes PPP Experience or No PPP Experience, since it depends on the year that each country has its financial closure.
A variable representing the total investment (investment in Source: Private participation in infrastructure projects database, World Bank.
facilities and in government assets) for each project was included(Total Investment). Its values are in 2005 constant US million dollars.
region, 0 otherwise. Asia becomes 1 if the project is in the South It is expected that a project needing more investment will have Asia region or in the East Asia and Pacific region, 0 otherwise. Latin greater difficulty achieving a positive outcome. Another variable America becomes 1 if the project is in the Latin America and the (Percentage Private) was set to show the percentage of the project Caribbean region, 0 otherwise. Projects executed in the Europe and company or project contract owned by private investors. The Central Asia region were taken as the base case and represented database projects may belong to one of the following transport when the three dummy variables became 0.
sectors: toll roads, seaports, airports, and railroads. One dummyvariable was created in order to report the type of sector in which the project belonged: Toll Roads became 1 if the project was a toll One dummy variable was created to classify whether by the project’s financial closure the country of the project was a low- orlower middle-income country or an upper middle-income country.
Low- and Lower Middle-Income variable became 1 if the country of The variable (Investors) was built in order to capture the effect of the project was a low- or lower middle-income country, 0 otherwise.
the number of private investors in a PPP project. illustratesthe frequency of the consortiums comprised of more than one private investor across the different regions. In general, 41% of the A dummy variable to include GDP growth was added (GDP total of PPP projects of the database involves more than one private growth). If, during the year of financial closure GDP growth of the investor, but these consortiums are primarily in Latin America and project’s country is negative, then the value of this dummy is zero.
If GDP growth is between 0% and less than 3%, it takes the valueone; if it is between 3% and less than 6%, it takes the value two; and it takes the value three if GDP growth is more than or equal to 6%.
The variable Number of Agencies was constructed to reflect the Another variable was included to measure the country’s develop- number of multilateral lenders in certain projects. As shown in ment: the current account balance as the percentage of GDP for a multilateral lender supported only 12% of the projects in each project in its year of financial closure (Account). Finally, in the database, and 57% of these are projects realized in Latin order to capture exogenous macroeconomic trends that might be affecting the results, the variable Trend was created, starting at 0 inyear 1984, and adding one for each year until 2005.
A 6-point scale variable Corruption was included for each country for the project’s year of financial closure. The value 6 was given to the most corrupted country as perceived during that year.
The types of corruption that the variable takes into account are Four models were estimated as shown in Model 1 was actual or potential corruption (excessive patronage, nepotism, job our first approach in modelling the hypotheses, where we focused reservations, loose ties between politics and business, etc).
on the effect of the variables representing countrys’ past experiencewith PPPs (Hypotheses 1a and 1b), macroeconomic performance 3.4.6. Country’s democratic accountability (Hypothesis 1c), corruption (Hypothesis 1d), democratic account- A 6-point scale variable Democratic Accountability was included ability (Hypothesis 1f), number of investors (Hypothesis 2a), and for each country for the project’s year of financial closure. The multilateral lenders (Hypothesis 3). As not all the variables were higher number of points is assigned if a country is closer to an statistically significant at a 95% of confidence, Model 2 was alternating democracy governance, while the lowest score is Incidence of multilateral support across the regions.
Three dummy variables were created to classify the region in which the project was executed. Africa becomes 1 if the project is in the Sub-Sahara Africa region or in the Middle East and North Africa Successful is understood as a project whose status was under construction, 9 Unsuccessful is understood as a project whose status was cancelled or Source: private participation in infrastructure projects database, World Bank.
P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 parameter for the variable Yes PPP Experience is statistically signif- icant and positive. This reflects the significance that past experi- ence in transport PPP projects plays in the success of futuretransport PPP projects. Past experience is not only a learning process, but also highlights a government’s reputation in honour- ing this type of agreement. In a similar way, Hypothesis 1b is strongly supported by the models, denoting a distinction between “good” and “bad” experience (successful and unsuccessful projects), and sanctioning failed past experience in transport PPP As expected, we also find a positive correlation between a country’s macroeconomic performance, reflected in the variables Account and GDP Growth, and the positive outcome of a PPP project (Hypothesis 1c). Both models acknowledge the importance of the variable Account. On the other hand, Models 1 and 3 indicate a positive impact of the variable GDP growth in the success of a PPP, proved to be significant in Model 3. The macroeconomic conditions on the models suggest the relevance of these indicators in pre- dicting the outcome of a PPP project. While good macroeconomic conditions may enhance the positive outcome of a PPP project, poor macroeconomic conditions may inhibit it.
In the case of the variable related with corruption, there is a negative association between countries perceived as more cor- rupted and successful PPP projects (Hypothesis 1d). This highlights the difficulties that PPP projects may face in more corrupted countries, where fewer investors are willing to undertake a PPP project, thus constraining the outcome of a PPP. Moreover, we can observe that, even if there are private investors willing to partici- pate in a PPP project, it could be possible that the selected private partners are the most inclined towards bribery or collusion with the political establishment rather than the most capable partners. The influence of corruption appears more prevalent in a project’s success if it is executed in Latin America and the Caribbean and Africa (Hypothesis 1e). These regions seem to be more sensitive to the perception of corruption, although the average of the country’s perception of corruption in Latin America is not the highest. This situation might reflect a market threat in countries perceived as Regarding Hypothesis 1f, both models show a strong positive relationship between developing countries perceived with low democratic accountability and PPP outcomes. Considering the countries in our dataset, this relationship highlights that a country Note: t statistics in parenthesis, * significant at 5 percent, **significant at 1 percent.
with a low democratic accountability score, perhaps an autarchy The authors used program R for computing models 1 and 2 and Biogeme ( government, may potentially have more authority to support the PPP project than a more democratic government. Infrastructureprojects (highways, ports, airports, etc.), which require large sunk estimated in order to fulfil this requirement. In Model 3 we wanted investments and a very long recouping period, if not fulfilled, may to upgrade the estimation by including the dummy variables rep- be seen as a major government failure. And this is particularly resenting the different regions of the world where projects are important in developing countries, where infrastructure invest- located (Hypothesis 1 h), and the interactions between the former ment may greatly improve quality of life. Governments with lower variables used and the regions (Hypothesis 1e), types of projects democratic accountability are more successful in PPP projects (Hypotheses 1 g and 1i), and the income level of the country because they seem more willing to fulfill the long-term require- (Hypothesis 2b). Model 4 resumes all of the hypotheses and shows ments, possibly in order to adhere to a populist approach. However, those that proved to be statistically significant.
the perception of democratic accountability seems to be more We find strong support for almost all of our hypotheses. In all relevant in all transport projects except toll roads, which is in line specifications the variable representing the total investment (in with the previous justification about the necessity for large capital facilities and government assets) proves to be significant. We find investments in these types of transport investments.
statistically-robust support for a negative association between the In order to use the dummy variables for regions (Hypothesis total investment and the success of a PPP project. This seems likely, 1 h), all the projects in Europe (and Central Asia) were regarded as as a higher total investment means a riskier project, which in turn the benchmark. As shown in Model 4, Asian (South Asia or East Asia makes it increasingly difficult to achieve a successful outcome.
and the Pacific), African (and Middle Eastern) and European Consistent with Hypothesis 1a, we observe a positive associa- countries bear the same risk in terms of transport PPP success.
tion between a country’s past experience with transport PPP Conversely, Latin American (and Caribbean) countries show a lower projects and the success of later PPPs. All the models show that the risk of failure. This could be due to the longer PPP experience that P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 most Latin American countries in the database have compared with or no experience and help them set up a regulatory and/or legis- other countries in other regions. Although projects from European countries have been more successful (in percentage) than Latin It is not surprising that GDP growth and the current account America’s, they are fewer in number and thus their PPP experience balance as a percentage of the GDP may impact on the success of a PPP project. Unfortunately, countries that require successful PPPs Turning to the hypotheses regarding investors (Hypotheses 2a often have very low (or even negative) GDP growth and a negative and 2b), we find enough evidence to support Hypothesis 2a. Vari- account balance. As also highlight, devel- able Investors proved to be significant to assert the importance of opment agencies should assist these countries to pull them out of the number of investors in a transport PPP project. As the number of investors increases, the chance of a successful PPP decreases.
The perception of a country’s level of corruption and democratic Larger numbers of private investors that form big conglomerates accountability appears to be relevant in the final outcome of a PPP may have increased difficulty in communication and a higher project. Countries with governments perceived as corrupted will chance of dispute among them. On the other hand, countries with hardly find international investors (often those with the most low- or lower middle-income appear to offset this result as the experience in this type of project). Projects developed in countries parameter representing this interaction appears to be positive and with governments perceived as having low democratic account- significant (Hypothesis 2b). These countries usually have lower ability can achieve better performance than projects in countries expertise in large infrastructure projects (and less in PPP projects), perceived as having higher democratic accountability. In this case it so greater investor expertise might prove to be more relevant than seems that autarchies may have a better capacity to assist PPP a communication problem. A project in a riskier country repre- projects, if needed, than in the case of alternating democracies.
sented by a low income status could, moreover, compel private The importance of the region where the project is located has investors to remain alert and involved in this particular investment.
proven to be relevant, making Latin American projects more Hypotheses 2c and 3 are not statistically validated by the models attractive for success, and thus for future investors. Although presented in The variable representing the existence of European and African projects in the developing world do not have multilateral lenders proves to be statistically insignificant, but its a poor record in terms of their success, they do have less experience positive sign confirmed at least that the suppositions described in PPP agreements in transport, and this situation could be previously were in the right direction. To understand these results, damaging their score (in relation to Latin American projects).
a correlation analysis was made and no indication of a correlation Development agencies should focus on these regions, not only to arose between these variables and the other ones modelled.
allow them to grow in terms of experience, but also to help them Previous results () have shown that before define a regulatory framework for PPP projects.
introducing such variables as corruption, democratic accountability A critical point in our research is certainly the definition used for and regions, these two variables were statistically significant, but the success of a PPP, since we consider a variable linked with their importance lessened and thus lowered their significance. As economic performance, rather than use a variable related to the shown in only 12% of the database projects had at least one status of a project. Our further research will be directed towards multilateral lender, so we will continue to analyze their importance obtaining more precise investment information in order to broaden as more projects (with more information) become available.
our results. We will compare the results with a similar analysis of In relation to the models, only Models 1 and 2 focus on the transport PPP projects in the developed world, since certain variables describing project and country, whereas Models 3 and 4 conclusions, such as the effect of corruption, may be different in this use the information provided by the first two models and add the scenario. Also, it would be interesting to study the success of PPPs interaction between variables and the region constants. As the log focusing within one transport subsector in order to add more specific likelihood increases, this information proves to be relevant for the characteristics and some efficiency indicators into the analysis.
estimation. The best model is Model 4, since it includes moreinformation about the variables and the interactions betweenthem; it is statistically superior than Model 3 (all its parameters are significantly different than zero); and, because a log likelihood- Bennett, J., & Iossa, E. (2006). Building and managing facilities for public services.
ratio test does not reject the null hypothesis that both models are Journal of Economics, 90, 2143e2160.
equivalent, for parsimony, Model 4 is better.
Bierlaire, M. (2003). BIOGEME: A free package for the estimation of discrete choice models. Proceedings of the 3rd Swiss Transportation Research Conference,Ascona, Switzerland.
Butkiewicz, J., & Yanikkaya, H. (2005). The effects of IMF and world bank lending on long-run economic growth: an empirical analysis. World Development, 33(33), PPP projects have gained relevance as a way to finance transport Dailami, M., & Klein, M. (1998). World Bank Policy Research Working Paper 1868.
infrastructure and services. PPPs have been supported by govern- Government support to private infrastructure projects in emerging markets.
ments, sponsored by the private sector, and have also been fav- Estache, A. (2004). PPI partnerships vs. PPI divorces in LCDs. World Bank Policy oured by multilateral agencies. Although there are numerous Research Working Paper 3470. January 2005.
Galilea, P., & Medda, F. (2007) Influence of foreign private investors and multilateral successful PPP projects, notwithstanding, there have also been lenders on the success of publiceprivate partnerships in transport. Second a large number of “divorces” (In this paper we have international conference on funding transport infrastructure, Leuven Belgium.
presented empirical evidence on the role that country experience Gould, D. J., & Amaro-Reyes, J. (1983). The effects of corruption on administrative in PPPs, private investors, and multilateral lenders may play in the performance. World Bank Staff Working Paper No. 580.
Greene, W. (2003). Econometric analysis. Upper Saddle River, New Jersey: Prentice Hall.
positive outcome of a PPP in transport.
Grout, P. (1997). The economics of the private finance initiative. Oxford Review of A country’s past experience in PPP agreements in transport is important, not only in attracting new investment projects, but also Hammami, M., Ruhashyankiko, J. F., & Yehoue, E. (2006). Determinants of pub- liceprivate partnerships in infrastructure. IMF Working Paper WP/06/99.
in instilling greater confidence in the success of present projects.
Hart, O. (2003). Incomplete contracts and public ownership: remarks, and an appli- This also means that countries with poor past experience, or no cation to publiceprivate partnerships. The Economic Journal, 113(March), 69e76.
past at all, will find it more problematical to complete successful Henisz, W., Holburn G., & Zelner B. (2005). Deinstitutionalization and institutional replacement: state-centered and neo-liberal in the global electricity supply PPP projects. However, if multilateral lenders want to promote PPP industry. American political science association conference, Washington, DC, investments, they should support projects in countries with limited P. Galilea, F. Medda / Research in Transportation Economics 30 (2010) 102e109 Huntington, S. P. (1968). Political order in changing societies. New Haven, CT: Yale Mauro, P. (1995). Corruption and growth. The Quarterly Journal of Economics, 110(3), Klitgaard, R. (1991). Gifts and bribes. In R. Zeckhauser (Ed.), Strategy and choice.
Medda, F., & Carbonaro, G. (2007). Publiceprivate partnerships in transportation: Some insights from the European experience. EIB Working Paper.
Laffont, J.-J. (2000). Incentives and political economy. Oxford: Oxford University Schleifer, A., & Vishny, R. W. (1993). Corruption. The Quarterly Journal of Economics, Laffont, J.-J., & Martimort, D. (2002). The theory of incentives-the principal agent Sirtaine, S., Pinglo, M. E., Guasch, J. L., & Foster, V. (2005). How profitable are model. Princeton, New Jersey: Princeton University Press.
infrastructure concessions in Latin America: empirical evidence and regulatory Laffont, J.-J., & Tirole, J. (1993). A theory of incentives in procurement and regulation.
implications. PublicePrivate Infrastructure Advisory Facility Trends and Policy Cambridge, Massachusetts: MIT Press.
Leff, N. (1964). Economic development through bureaucratic corruption. American United Nations. (1989). Corruption in government. New York: United Nations.
Valila, T. (2005). How expensive are cost savings? On the economics of pub- Martimort, D., & Pouyet, J. (2006). Build it or not: Normative and positive theories of liceprivate partnerships. EIB Papers, 10(1), 95e119.
publiceprivate partnerships. Centre for Economic Policy Research. Discussion World Bank Independent Evaluation Group. (2007). A decade of action in transport.

Source: http://www.brt.cl/wp-content/uploads/2011/01/Galilea-P-Medda-F.pdf