How innovative Was the Poupatempo Experience in Brazil? institutional Bypass as a New Form of institutional Change

This article analyzes Poupatempo, a recent bureaucratic reform in Brazil. We argue that it offers insight into a potentially successful strategy to promote institutional reforms. After analyzing the design of the project and the obstacles to its implementation, we conclude that Poupatempo’s most innovative feature was the fact that it did not try to reform existing institutions. Instead it created a new pathway around an inefficient bureaucracy. This type of reform — the institutional bypass — has not been considered by the academic literature on institutional change but, as the case of Poupatempo suggests, it should be further investigated and analyzed.


introduction
A t a very intuitive level, everybody can agree with the idea that dysfunctional public institutions are undesirable.Nobody wants to deal with inefficient, incompetent and corrupt bureaucracies, courts or elected bodies.Yet, there is at least one other important reason for countries to want to get rid of dysfunctional institutions: they may impact on a country's development prospects.New institutional economists argue that developing and transitional economies struggle because their institutions 1 create incentives that are bpsr (2011) 5 (1)  12 11 -34 How innovative Was the Poupatempo Experience in Brazil?
ill-suited for economic growth: "Third World countries are poor because the institutional constraints define a set of payoffs to political/economic activity that does not encourage productive activity" (North 1990, 110).In other words, for new institutional economists having functional institutions may positively influence a country's development prospects.
A number of studies have now shown that not only is there empirical basis to support North's argument (that good institutions promote economic growth) but also that good institutions can have a positive impact on literacy levels and life expectancy (Rodrik et al. 2004;Kaufmann et al. 1999;Kaufmann 2003;2004;World Bank 2007).Thus, functional institutions are of utmost importance to any country, but they are especially relevant for developing countries such as Brazil.
Why then is Brazil not getting rid of all dysfunctional institutions at once, firing inefficient bureaucrats, prosecuting corrupt civil servants and creating public bodies that will allow it to develop at an even faster pace?The answer seems to lie in the numerous obstacles to institutional change identified in the academic literature, especially by new institutional economics (NIE).There are particular conditions and circumstances that ideally need to be present for institutional change and reform to take place, but these cannot be easily found.Instead, one often finds obstacles to institutional change, for example, strong political resistance to reforms, or informal rules and norms that reinforce the status quo (Trebilcock and Daniels 2008;Mahoney and Thelen 2010;North 1990;2005).This is not to say that institutional change never takes place or that it cannot be explained by NIE.On the contrary, institutional change has a prominent position in political science and new institutional economics has become increasingly relevant as a method to analyze these changes (Hall and Taylor 1996;Rezende 2009).For example, NIE has been used in political science to analyze constitutional change, administrative reforms, welfare state reforms, sectoral reforms, and fiscal reforms, among others. 2  Some of these studies have proven to be illuminating in helping us understand how institutions emerge, evolve and change, but there are limitations related to the theoretical assumptions of this type of analysis (Rezende 2009).Cognizant of that, this article suggests that even if one adopts its theoretical assumptions, there is yet another important limitation: NIE's focus on changes to pre-existing institutions.
The literature on institutional change has focused primarily on attempts to reform a dysfunctional institution and the obstacles that reformers may face in this process, including fierce and powerful resistance of interest groups that benefit from the status quo and pathdependency.This article acknowledges that this is an important focus of analysis, but it argues that it does not need to be the only one.We argue that one could (and should) also focus on attempts to promote institutional change by bypassing dysfunctional institutions, instead of modifying them.This is what we call an institutional bypass, a type of institutional bpsr (2011) 5 (1) 13 11 -34 Mariana Mota Prado and Ana Carolina Chasin reform that creates new pathways around clogged or blocked institutions (Prado 2011).
To illustrate the importance of this alternative focus -institutional change that occurs outside pre-existing institutions -this article analyzes a recent institutional reform in Brazil called Poupatempo (in English, "timesaver").Poupatempo is a one-stop shop for Brazilian citizens who need documents such as identity cards, driver's licenses, and criminal records.
Public utility companies also have offices at the Poupatempo units, where consumers can pay utility bills (electricity, water and telephone), dispute charges, and settle debts with representatives of these companies.In contrast to the pre-existing bureaucracy, which has offices scattered throughout the city, Poupatempo places offices of the federal, state and, in some cases, local administration in one location of easy access to the general public (normally in the vicinity of subway and bus stations).This makes it easier for citizens to obtain a wide range of services.Moreover, in some cases the services at Poupatempo can take significantly less time than the ones offered by the pre-existing bureaucracy.
The project started in 1997 with one unit in the capital and today it has a total of 28 units in the state.In 2010 alone the units combined provided services to more than 32 million people. 3In 2011, Poupatempo was providing services to an average of 115,000 people a day. 4 The impact of the Poupatempo is significant: the total number of people who benefited from services provided by Poupatempo from 1997 to 2010 is approximately 272 million (Poupatempo 2011).We argue that this fast expansion and ability to overcome obstacles to institutional reforms can be largely attributed to the fact that it was an institutional bypass.In other words, it did not face the same obstacles related to path-dependency and transaction costs that other attempts to reform face.
To support our argument, this article analyzes the implementation of the Poupatempo project identifying the obstacles it faced.For this purpose we have used the scarce academic literature available, complementing it with fieldwork, which consisted of analyzing official governmental publications, visiting some units and interviewing some of the relevant actors in the process.We first gathered all the public information available on the web-site and official publications and, after examining it, began to schedule the interviews.Two kinds of interviews were conducted: interviews with the individuals responsible for creating and initiating the implementation of the Poupatempo project; and interviews with the staff currently responsible for its administration.In the first group, we interviewed Daniel de Sequeira, the director of the identification department.All of them were semi-structured interviews which had the purpose of both furnishing some institutional data, and presenting the individuals' understanding and own thoughts about the program.These interviews were used as primary sources of information, providing data and details that were not readily available in official documents or academic scholarship.
The article will be structured as follows.The first part maps out the conception of the project and identifies its constitutive elements, trying to identify what made Poupatempo more efficient than the pre-existing state bureaucracy by analyzing its core innovations.In the second part, we discuss implementation, demonstrating how Poupatempo overcame wellknown obstacles to institutional reforms, such as lack of resources, embedded cultural traits, political economy problems, and institutional interconnections.Our main argument is that Poupatempo managed to overcome these obstacles because it was an institutional bypass.
The third part asks whether Poupatempo can be considered a success, concluding that there are no conclusive answers to this question.The potential effects that an institutional bypass may have on pre-existing institutions are especially unclear.We suggest that Poupatempo may have had feedback effects on the traditional bureaucracy, promoting structural changes in the internal structures of the Brazilian State.This needs to be further investigated in future research, in order to determine whether institutional bypasses like Poupatempo could be an important drive for institutional change.In any event, our conclusion is that regardless of its impact on pre-existing institutions, Poupatempo is clearly an institutional bypass, which is a concept that deserves more attention by the specialized literature.

Poupatempo: A Brief Overview of the Project
One-stop shops for bureaucratic services have enjoyed a significant boom in the last decade around the world (Programa Nacional de Gestão Pública e Desburocratização 2009), and these projects could serve as interesting case studies on successful attempts to promote bureaucratic modernization by tailoring private sector management techniques to the public sector, a strategy known as New Public Management (Israel 1987;Manning 2001, 298-99).
The one-stop shop idea, for instance, was influenced by the business strategies of shopping centres and supermarkets (Estrada 1982).In Brazil, previous reforms had focused on traditional public sector management principles, such as hierarchical structures, anonymity of public servants, and input and output centred program management.In contrast, Poupatempo's staff asserted that this project centred on the customer, deemphasized hierarchy, and focused on outcomes (Gonçalves 2003, 9).This shift in the focus and approach of bureaucratic reforms could serve as a potential explanation to Poupatempo's success and could provide lessons for bureaucratic reforms in other countries.Thus, Poupatempo could be interpreted and analyzed as a relevant case of public management and public administration reforms.However, we believe that Poupatempo can also offer insights that address broader concerns about institutional reforms, which are relevant to but not confined to the bureaucratic sphere.The latter will be our focus in this article.More specifically, Poupatempo may show an effective strategy to overcome obstacles to institutional reforms and institutional change identified in the literature.However, before we get to these broader institutional concerns, it is important to explain the context in which Poupatempo was created, and what kind of innovations it has implemented in an attempt to achieve a higher level of efficiency than the traditional bureaucracy.In other words, this first part of the article will provide relevant background information about Poupatempo, so as to help the reader better understand the obstacles it faced -or could have faced -in its implementation.More specifically, this section will show that while it is not clear where the idea of a one-stop shop for bureaucratic services was conceived, there was strong political will to promote reforms within the office of the governor.In terms of innovations, we show that Poupatempo implemented both creative and simple solutions to problems that were impairing efficient delivery of services.

the unknown origins of the idea of a one-stop shop for bureaucratic services
Within Brazil, it is hard to identify where or when the idea of a one-stop shop for bureaucratic services was originally conceived.While the state of São Paulo was designing its project, a similar project had already been implemented in 1995 in Bahia: Serviço de Atendimento ao Cidadão ("Citizen Assistance Service").The experience in Bahia influenced the creation of the Poupatempo, but its creators often cite significant differences between these projects to support the idea that the latter did not merely mimic the former (Annenberg, Tomchinsky and Tokairim 2006, 29).They argue, instead, that the proposal of a one-stop shop for bureaucratic services was originally presented in a book (Estrada 1982) similar reforms at about the same time.For instance, in the same year the Poupatempo was created, 1997, Australia implemented a one-stop shop agency, Centrelink (Halligan 2005, 134).Portugal and Spain have also adopted the one-stop shop model for bureaucratic services (Programa Nacional de Gestão Pública e Desburocratização 2009).
It is also not fully clear what mechanism led to the proliferation of this type of bureaucratic reform in Brazil and around the world in the 1990s.One hypothesis is that it was part of a larger process to increase bureaucratic efficiency led by multilateral institutions such as the World Bank.Another hypothesis is that these projects were simply the result of a process of policy diffusion, i.e., informal exchanges of information among public administrators about successes that could be replicated in other places.For instance, between 1997 and 2003, Bahia received 22 delegations from other countries interested in replicating the experience, as well as receiving requests from other Brazilian states to provide technical assistance in the creation of their own one-stop shop units (Gonçalves 2003, 11).It is likely that is was a combination of both, as after an initial process of informal diffusion, the reform was endorsed by governmental and international institutions: the Inter-American Development Bank sent a mission to Bahia in 1996; in 1997, the United Nations recommended the replication of this type of reform in all continents; and in 1998, the Brazilian government began supporting similar reforms in other states (Gonçalves 2003, 13), and in 2002 it published a manual 5 on how to implement the bureaucratic one-stop shop (Ministério do Planejamento, Orçamento e Gestão 2002).This national and international endorsement may not only have provided funding and political support, but it may have created incentives for a more intensive exchange at the level of informal networks.
Although it is hard to pinpoint where the inspiration for the one-stop shop model came from, or where it was originally conceived, how Poupatempo was created is clear.The project was an initiative of the executive branch of the state of São Paulo (Paulics 2003).
In 1994, Mario Covas was elected governor of the state of São Paulo with an agenda of improving the provision of services to citizens and reforming the public administration.
In January 1995, when he was inaugurated, he created the Unit of Strategic Management, which was in charge of designing a plan to execute the governor's campaign promises. 6  One year later, this unit proposed a project entitled Central de Atendimento à População ('Citizens Services Centre'), which later changed its name to Poupatempo (Annenberg et al. 2006. 29-30).
implementing the idea: Diagnosing the problems and designing tailored solutions for each office of the bureaucracy In order to implement the Poupatempo project, in September 1996 the governor of the state of São Paulo assembled a team of professionals with diverse backgrounds.This bpsr (2011) 5 (1)  17  11 -34 Mariana Mota Prado and Ana Carolina Chasin team was comprised mainly of civil servants from different governmental departments, but it also included some political appointees and private sector professionals. 7The team was in charge of designing one unit, which later became a model for new ones.The coordinator of the project at the time points out that the task was experimental: the team had not done anything similar before, and at the time it was unknown how the project would unfold (interview with Daniel Annenberg, 2009).The team started with a case-by-case diagnosis of the problems in primary government services, and based on this decided which problems would be tackled by the project (Paulics 2003, 6).Some government offices did not need improvements and became part of Poupatempo right away, with no changes in their existing processes.One example is the public utility company for water services (Sabesp), which had an efficient system of consumer services. 8In contrast, other offices and departments required more careful analysis of why service times were long, and whether improvements could be made to expedite their delivery.An example is the State Identification Department, which formerly took up to 60 days to issue identity cards (Registro Geral, also known as RG) for residents of the state of São Paulo. 9In this case, the Poupatempo team made a diagnosis and prepared proposals in collaboration with the offices of the pre-existing bureaucracy (interview with Daniel Annenberg, 2009).
Some diagnoses made by the team were fairly simple.For instance, in the case of the national identification card, the pre-existing bureaucracy was using paper files: citizens would fill out a form that would be typed into a paper file and then shipped to a central office, with the individual's fingerprints.Once the fingerprint file was in place at the central office, the offices of the bureaucracy would be authorized to produce a new identity card.
To replace a lost or stolen identity card, new fingerprints would be taken.These would be shipped to the central office to be compared to those already on file.If there was a match, the office of the bureaucracy would then be authorized by the central office to issue the replacement card.The time required to physically transport files from local offices in different locations around the city and to store these documents was substantial, and it resulted in identity cards being delivered to citizens around 60 days after the date in which the form was filled.
The team's proposal, in this case, was to install a scanner that would transfer the file to the central office electronically, reducing the required time to physically transport and store the files.As the director of the Registration Department noted, a scanner was installed when the first unit of the Poupatempo was created, significantly reducing the time for the cards to be issued (interview with Carlos Sequeira, 2009).This example illustrates two principles that guided the project.First, the reform was oriented to immediate results: the team avoided creating a costly digitalization project that would take many years to be implemented.Instead, the Poupatempo office had a scanner at the time of its creation but the bpsr (2011) 5 (1)  18 11 -34 How innovative Was the Poupatempo Experience in Brazil?
pre-existing bureaucracy continued to operate with paper files.Second, the project focused on the "consumer" or the user of the service, not on the structure of the bureaucracy.In other words, it did not try to change hierarchies and the functions performed by civil servants.

Poupatempo's strategies to increase efficiency and improve quality in the delivery of bureaucratic services
Poupatempo's goal was to guarantee access to basic governmental services with expediency, quality and transparency.To achieve this goal, the project used a series of strategies. 10 Technology was used to improve the flow of the service and to eliminate bottlenecks caused by inefficient or redundant processes.Hardware played an important role.As the example discussed earlier showed, adopting the scanner eliminated the need to physically transport the files.Overtime, a new digitalized system allowed the complete substitution of paper files with electronic files.As a result of the digitization process, issuing identification cards became faster, cheaper and more secure (Annenberg et al. 2006, 148-49).
The use of hardware (such as scanners) was coupled with sophisticated information technology, which provided for a more efficient flow of information.For instance, Poupatempo had a tracking system for a customer's service at the reception desk.The system measured how long a person had to wait in line, how long the line was, how long the person was in contact with customer services, how productive each employee was etc.This system allowed the manager to allocate more resources (such as personnel) to the services in greater demand.While this system managed information regarding the interaction between citizens and offices, there was another system to manage the flow of information between the offices.and citizens would be required to make three or fours trips to the same office just to place a request for services.In addition to the booklets, Poupatempo also has a call centre providing information about the services available on each unit and a list of the required documents for each service.In 2002, the attempt to reduce delays even further led to the creation of the internet service called e-Poupatempo. 11 Another concern of the reforms was appropriate infrastructure.The building structures of Poupatempo units were designed to be easily accessible and comfortable for those waiting for the service to be provided.When choosing the location to install a future unit, the architects would look for places situated in central areas, easily accessible by public transportation.They also sought large ventilated buildings, with high ceilings and natural sun light.Once a building was acquired or rented, they would divide areas of circulation from service areas, and each service area would be identified by different wall colors and colored signs.Finally, there are no walls dividing different departments and offices.Instead, the units have half walls in order to give the feeling of openness and transparency (Annenberg et al. 2006, 123-45).
Regarding human resources, an internal selection process was able to identify civil servants who were willing to be transferred to a Poupatempo unit.They were selected through a competitive internal process, which evaluated, among other things, their capacity to interact with people, to take initiative when needed, and to be flexible and open to change.The incentive to participating in this process was a marginal salary increase, a better working place and a more dynamic environment. 12After the selection process, these servants underwent intensive training to learn how to operate the new system, and once they became employees of Poupatempo, they were periodically re-trained.
This internal selection process, targeted at motivated civil servants, was used in 1997 and 1998, but was largely discontinued in 1999 due to complaints of the heads of offices in the pre-existing bureaucracy, where employees were being selected and removed.
They claimed that Poupatempo was draining their core personnel, leaving the older departments with fewer employees than they needed (and a hiring freeze did not allow them to hire replacements).The former adjunct secretary reports that, as a result, around Overcoming Obstacles to reforms: Poupatempo as an institutional Bypass Theories of institutional change have often identified significant obstacles to reform.
First and foremost, political will is a necessary condition for institutional change to take place (Mahoney and Thelen 2010).This was certainly present in the case of Poupatempo, as we have shown in the previous section.While necessary, political will is not a sufficient condition to promote reforms.Reforms need to be politically desirable, but they also need to be politically feasible.Thus, if there is resistance or obstacles to the implementation of institutional reforms, it is unlikely that they will succeed.
There are three main obstacles to institutional reforms, according to Trebilcock and Daniels (2008).First, despite the political will to promote reforms, countries may lack the necessary financial, technological or human resources to implement changes.Second, a series of social values, norms, attitudes and practices, which Trebilcock and Daniels loosely classify as social-cultural-historical factors, may form a hostile environment for implementing reforms.Third, there are political economy-based impediments.Interest groups will resist reforms that eliminate their privileges, do not foster their interests, or do not offer any gains (material or otherwise).The last two obstacles can potentially explain why the massive surge in development assistance for institutional reform projects in developing and transition economies have had mixed to disappointing results thus far: there was money to tackle the lack of resources (Trubek 2006, 74), but the other two problems were not properly dealt with (Trebilcock and Daniels 2008;Tamanaha 2011).
In contrast to those identifying obstacles that explain why institutional reform fails to take place or to succeed, Mahoney and Thelen (2010) have identified, in their most recent book, what factors explain institutional change.Not surprisingly, their analysis largely matches two of the obstacles to reform identified above.First, they argue that the political context matters, i.e., if the group that is likely to defend the status quo is not powerful and their veto power is weak reforms are likely to happen.This largely maps into the politicaleconomy problems described above.Second, they argue that the characteristics of the institutions also affect the likelihood of reform.They distinguish between institutions with high or low level of discretion in applying rules, and argue that lower levels of discretion are more conducive to reforms.This is largely because informal rules can undermine efforts to promote reforms, as those who are interpreting, applying and enforcing the new rules are Mariana Mota Prado and Ana Carolina Chasin not the ones that made them.Thus, the more discretion they have in applying the rules, the more they can deviate from the purpose of the reform.This largely maps into the socialcultural-historical factors described above.
The fact that Poupatempo was successfully implemented and expanded significantly within a relatively short period of time suggests that the establishment of Poupatempo' was able to overcome the three obstacles to institutional change mentioned earlier, namely: lack of resources, culture (embedded patterns of behaviour or values that negatively impact reform efforts), and political economy (groups of interest that benefit from the status quo resist changes for self-interested reasons) (Trebilcock and Daniels 2008).Moreover, the reform also needed to tackle the problem of institutional interconnections, where other preexisting institutions (in addition to the one being bypassed) can undermine the effectiveness of reforms due to self-reinforcing mechanisms and institutional interdependencies (Prado and Trebilcock 2009).
Considering that these obstacles are likely to be present in most attempts to promote institutional reforms, how was Poupatempo able to overcome them?We argue that the reason is because Poupatempo was an institutional bypass.An institutional bypass has four characteristics: (i) it keeps pre-existing institutions in place; (ii) it creates an alternative pathway creating the option to use one or the other; (iii) it aims at being more efficient or functional than the pre-existing institution and (iv) it only offers some services or performs just some functions that used to be provided or performed by the pre-existing institution, and therefore it does not modify the entire system (Prado 2011).
Poupatempo exhibits these four characteristics.First, the offices of the old bureaucracy were not eliminated or replaced at the time of Poupatempo's creation.As we will discuss below, some of the offices later faded away and died, but this was a consequence of the success of Poupatempo, rather than an objective of the project.Second, citizens were given the option to obtain documents in the old bureaucracy, or in the new Poupatempo.Third, due to a series of innovations described above, Poupatempo aimed at (and in certain cases succeeded in) significantly improving the provision of documents by reducing the costs and time required to obtain them.Fourth, Poupatempo did not replace all the functions of the pre-existing bureaucracy.For instance, Poupatempo performed two functions of the department of motor vehicles (issuance and renewal of driver's licenses and registration of vehicles), but it left all the other functions to be performed by the pre-existing bureaucracy.Finally, as an institutional bypass, Poupatempo was able to overcome the major obstacles for the implementation of institutional reforms in developing countries mentioned above: lack of resources, culture, political economy and path dependence.The following sections will discuss whether and how Poupatempo overcame each of these obstacles.Obstacle 1: resources Poupatempo was able to overcome lack of resources by starting with a small pilot project that did not require a significant initial investment.Indeed, the initial budget for the first unit of Poupatempo project was approximately R$ 10 million (US$ 5 millions). 14In 2008, it its annual budget had increased to R$ 150 million (roughly US$ 75 millions) a year.
In large part, the budget increase was a result of the success of the pilot project, illustrated by the increasing demand for the services.Soon after its inauguration in 1997, the number of people looking for the services quickly became larger than the unit's capacity.According to the team's manager "it had more success than it could handle."The state secretary of government and public management reports that, as the project gained popular support, it was easier to convince the Office of the governor that it was a good investment, making it attractive to use part of the budget to expand it.The governor's office monitored numbers and statistics after the implementation of the project, and the demand in the first unit was a key factor in convincing the governor and other state officials of the need to invest in and to expand the project (interview with Antonio Angarita, 2009).
Also, as the project began displaying substantive results, it became much easier to overcome political resistance towards allocating resources to the project.Popular demand made Poupatempo an attractive project for electoral purposes.Voters were the ones benefiting from its services, and politicians began to see Poupatempo as a very attractive "campaign card".After the initial success of the first unit, a series of units were opened in the capital and in other cities by the end of 1998, many shortly before state elections.
Increasing the Poupatempo budget, however, was not as easy as it might seem.The state government has a complicated budgetary process in which what is defined as budget for one year is replicated in the next year.Therefore, adding a brand new project in the budget was a difficult process, because it meant taking money out of other items, which was always met with fierce resistance.To overcome this difficulty at the early stages, the governor's office presented Poupatempo as a temporary project entitled to exceptional funding (crédito suplementar).As the former adjunct secretary reports, eventually, the success of the project, created enough political support to have it included in the budget (interview with Dalmo Nogueira, 2009).
One budgetary concern with the project, however, was the duplication of bureaucratic services.Multiple offices serving the same function seems irrational from the point of view of resource expenditures.So, while the success of Poupatempo allowed for the expansion of its budget, this expansion was always considered within a context in which redundancy would be a waste of money.This resulted in the eventual elimination of some of the offices of the pre-existing bureaucracy, like the provision of the identity card services, which are now exclusively offered at Poupatempo.A key aspect of cultural problems in bureaucratic reforms is institutional culture.
Culture here refers to informal rules and institutional practices that are difficult to change, in part because they are self-reinforcing (North 2005;Trebilcock and Daniels 2008;Prado and Trebilcock 2009).Poupatempo overcame obstacles related to institutional culture in at least two ways.First, it adopted a recruiting process that tried to screen out candidates to eliminate patterns of behaviour that would reproduce the problems of the pre-existing bureaucracy.At first, the team considered that for this purpose hiring new people would be better than recruiting civil servants from the pre-existing bureaucracy.This option, however, was not favoured by the governor, who believed that there was a need to change the mindset of the existing bureaucracy (interview with Dalmo Nogueira, 2009).In addition, those involved with the reform process noted that hiring new people was not feasible at that time because at both the federal and state levels, there were massive efforts to reduce the size of bureaucracy (which was part of a larger plan to promote fiscal stability) (interview with Antonio Angarita, 2009).
Second, there were constant adjustments and improvements largely based on user's feedback and external evaluations.This reflected an understanding by the reformers that efforts to modify embedded cultural traits can result in consequences that are hard to identify in advance (Prado and Trebilcock 2009).In this regard, it was extremely relevant for Poupatempo to have both specific suggestions from users, who were invited to provide feedback on the spot by filling out a form, and also from outsiders hired to provide external assessments of the overall performance of the units. 15  It is not clear, however, if these practices of constant adjustments and revisions that happened intensively at the beginning will remain in place.They may not, for instance, if the state government's urgency to introduce new units as fast as possible (in order to realize the electoral benefits associated with that) takes the improvement of the actual services off the government's priority list.The fast expansion of the project would then have a negative impact in the overall quality of the services offered.These practices of consulting citizens and asking for feedback may also disappear if the governor's office changes hands, which may be followed by the replacement of top personnel in the bureaucracy.The system seems to have been solid and consistent until now largely because the same party has been in power since its creation.Poupatempo's success may keep it in place, but there is a risk that another party may not want to have its name and its administration associated with a project created by the opposition.
Feedback has another beneficial role: it creates a self-reinforcing system of changes and constant improvements, modifying not only the culture of the services themselves but the culture of the people using the services.In this regard, one of the goals of the project was to prove that "the old image of inefficient and dysfunctional bureaucracy could be eliminated and overcome" (Annenberg 2002).The idea of asking for feedback may create a culture in which citizens feel entitled to demand better services, and to request changes.
For this reason alone, it would be relevant to secure its maintenance.
Obstacle 3: Political economy problems 16   Political economy problems is an expression that refers to obstacles created by entrenched interests.The reformers implementing Poupatempo had to adopt a series of strategies to overcome the resistance offered by certain groups at different phases of the implementation process.
The governor was able to overcome any legislative resistance by creating a project that did not depend on legislative approval.The governor had been elected with two major reform plans: privatization and internal restructuring of the bureaucracy.Both aimed at increasing efficiency (Banco Nacional de Desenvolvimento Econômico e Social 2002).
However, both required legislative approval and the governor thought it was too risky to put forward two major reform plans before the legislature at the same time.He then decided to present to the legislature only the privatization plan, abandoning the idea of promoting an internal restructuring of the bureaucracy, which was largely aimed at reducing the number of departments subordinated to the governor's office.Its closest advisors were left with the challenge of coming up with another plan to increase bureaucratic efficiency for services that would not be privatized, without internal restructuring.As the former state secretary reports, Poupatempo was an attractive solution because it did not depend on legislative approval.The governor could create and implement the project by decree, bypassing any potential resistance or obstacle that could be imposed by the state legislature (interview with Antonio Angarita, 2009).
However, despite bypassing legislative resistance, the implementation of Poupatempo still came up against a number of political economy issues.In some cases, there was initial resistance by certain groups inside the bureaucracy.Civil servants resisted the creation of Poupatempo largely due to rent seeking, i.e., when bureaucrats stood to lose some benefit (financial or otherwise) from having services transferred to Poupatempo.
Indeed, the creation of the project triggered resistance from corrupt civil servants who either received bribes to expedite the process to issue documents, or were making money by selling falsified documents.In these cases, the governor dealt with matters with an iron fist: he removed bureaucrats resisting reforms from their positions, or transferred them to another department of lesser importance.The former adjunct secretary mentioned that this happened, for instance, with sheriffs involved in corruption schemes who did not want the issuance of identity cards to be transferred from police stations to Poupatempo (interview bpsr (2011) 5 (1)  25 11 -34

Mariana Mota Prado and Ana Carolina Chasin
with Dalmo Nogueira, 2009).The governor's strong commitment to the project helped overcome some of these obstacles.This not only suggests that political will is important to implement reforms, but also that institutional bypasses may not be shielded from political resistance.In any event, one may be tempted to suggest that resistance could have been fiercer if reforms were internal (i.e., if they were not a bypass), but it would be hard to find evidence to support this counter-factual argument.
Nevertheless, the governor's power was limited to the state bureaucracy.He could not remove or relocate bureaucrats at the federal or the local level of the bureaucracy.In these cases, bureaucratic resistance was not so easily overcome.For instance, the former adjunct secretary mentioned that at the federal level, the federal police did not agree to have an office to issue passports at Poupatempo units (interview with Dalmo Nogueira, 2009).They claimed that Poupatempo standards of expediency would compromise the Federal Police security standards, but one can perhaps wonder if they were afraid of losing the monopoly over the services; or if they were resisting the idea of working longer hours, since the Poupatempo units had extended hours of operation.This is not to mention the possible interest to keep rents derived from corruption.
At the level of the local bureaucracy, there was also resistance, but largely motivated by a different set of problems.Local bureaucracies had very low standards of service provision, and they would require significant reforms before reaching the standards that would make it possible to consider their inclusion in the Poupatempo project.However, most municipalities could not afford these reforms.One of the few exceptions is the municipality of São José dos Campos, which is very wealthy.In addition, the mayor of the municipality was a political ally of the governor, which, as the former adjunct secretary notes, also helped create the goodwill to move ahead with the project (interview with Dalmo Nogueira, 2009).The former state secretary reports that, since reformers would need support from mayors to include local services in the Poupatempo units, local services were offered only in the cities in which mayors were political allies of the governor (interview with Antonio Angarita, 2009).Some civil servants who did not have self-interested reasons to resist the reforms were also not ready to support them.
The reasons for this resistance are not fully clear, but the uncertainties associated with the outcomes of the reforms, due to its experimental nature, may have played a role.
The superintendent reports that, in order to overcome this resistance, the reformers used meetings and a long consultative process, in which bureaucrats in the pre-existing bureaucracy were invited to voice their concerns and suggestions.In order to diagnose the problems of each department or office of the government, the Poupatempo gathered information about service provision in that department and then articulated concerns in meetings with representatives from each government office.The team used these meetings and the information produced by the assessments to convince bureaucrats of the importance of the project (interview with Daniel Annenberg, 2009).
In addition to the pre-existing bureaucracy, the government also faced resistance from other interest groups.In particular, there was significant resistance from business professionals who offered services connected with bureaucratic services.The state secretary mentions that, for instance, doctors who provided medical examinations for driver's licenses lobbied against offering them at the Poupatempo units because that would imply that civil servant doctors, working for the Brazilian public health system, would perform the examination in loco (expediting the services, and reducing the time required from citizens to get their licenses issued).He also describes another example: the middlemen with personal networks in the bureaucracy that offered expedited services for a fee (despachantes), often sharing a percentage of the fee paid with the bureaucrat who processed the paper work.
Citizens who were able to pay this fee would have their application processed faster through the back door.These groups strongly lobbied against Poupatempo issuing driver's licenses, as it could potentially kill their main (and perhaps only) source of income.Despite this resistance, the service was implemented in all Poupatempo units (interview with Antonio Angarita, 2009).

Obstacle 4: institutional interconnections
One important obstacle to institutional change and reforms is institutional interconnections.The former adjunct secretary explained that in the case of Poupatempo, this problem arose when the reformers were deciding who would control and coordinate the project.Poupatempo needed to be housed by an institution within the state government (interview with Dalmo Nogueira, 2009).Housing it in the office of the governor's chief of staff would guarantee that Poupatempo would be under direct control of the governor's office.However, according to the reformers, the bureaucratic structure of this particular office was ossified, and it would be difficult to change practices and procedures to increase efficiency and effectiveness in the decision making process despite the governors' strong commitment to the reform (interview with Antonio Angarita, 2009).
Take, for instance, budgetary allocations.If the project was housed within the chief of staff's office, the project's budget would need to be approved yearly by the state legislature.
In addition, the office of the governor must use public bids for every service or good provided by private companies.For instance, hiring architects to design and build new units could easily take six months.For this reason, the team reports that they decided to house the project in a state-owned company in charge of information technology and data processing (Prodesp) (interview with Antonio Angarita, 2009).This company had more flexibility in managing their budget.They could, for instance, hire an architect within a day (interview with Dalmo Nogueira, 2009).The downside of housing it at Prodesp was that Poupatempo could not be included as a formal item in the government's budget, making its budgetary situation very precarious, with no guarantee of funding from year to year.This shows how institutional interconnections and interdependencies are a relevant obstacle to reform, and also how an institutional bypass has more flexibility to avoid problems than reforms to pre-existing institutions would have.

Promoting institutional Change: the Consequences of Poupatempo
Can we describe the Poupatempo experience as successful?Yes, if we are simply concerned with institutional change.Poupatempo did promote significant bureaucratic reforms in Brazil.The answer may also be yes, if we consider that Poupatempo increased the overall satisfaction of citizens using the state bureaucracy, and gathered enough public support to allow for a significant expansion of the service in a short period of time.In this regard, one positive outcome of Poupatempo concerns effectiveness.Identity cards, which are now issues in 5 days instead of 60, are frequently mentioned as an example of the efficiency of the project.
However, there are a number of potential problems with the project.First, it is unclear how much more some of the services are costing the government under the reforms compared to providing the services prior to reform.One independent study concluded that the issuance of documents under the Poupatempo system seems to be on average more expensive for the state (R$ 16 in contrast to R$ 12 through the traditional bureaucratic procedure), but, on the other hand, the costs for the population to obtain these documents is much lower (R$ 17 in the Poupatempo, against R$ 34 in the traditional procedure) (Ferrer n.d.).However, as the superintendent affirms, there is very little information about the cost of the services pre-Poupatempo (interview with Daniel Annenberg, 2009).Therefore, the accuracy of this study might be questionable.Further research is needed to develop a cost-benefit analysis of the reforms discussed here.
Another problematic aspect of the project is the fact that it largely duplicates a series of services, by offering them both at the pre-existing bureaucracy and at the Poupatempo level.This entails start-up costs (building the unit, buying new equipment, furniture, hiring new personnel etc) and the total budget allocated for the delivery of a particular service needs to increase as a result.This initial investment occurs only once and permanent costs will be only those for maintenance and salaries, which need to be considered in light of the increased capacity in service provision and the circumvention of inefficiencies in the pre-existing bureaucracy.Nevertheless, there were claims that sometimes the project was increasing rather than reducing government's expenditures.It seems that in some cases this This is what happened with the issuance of identification cards.However, further research is needed to make a cost-benefit analysis of this change, and determine whether it was in fact beneficial to the state.More specifically, it is hard to determine whether the costs for the state were reduced in this case, as indicated above.Moreover, these cards were originally issued in police stations.Therefore, the costs to maintain the stations and its personnel remain, what may indicate that the overall cost for the government may be significantly higher now.
Another example in which the original offices were closed -and costs may have been reduced -is the consumer protection services (PROCON).This service is offered by the Consumer's Protection and Defense Foundation and, in the case of São Paulo, is linked to the State Department of Justice and Citizenship Protection.PROCON handles consumer's complaints against companies and service providers, using mediation and bringing the issue to court if necessary.The offices in which these services were provided were closed when the PROCON offices at Poupatempo units began to respond to a significant number of consumer complaints.In this case, the reduction in costs may be clearer, but further research is still necessary to assert this conclusively.
Another aspect of Poupatempo that may deserve further research is the potential elimination of pre-existing institutions in the future, as the project progressively becomes citizens´ first option for bureaucratic services.Indeed, an opinion poll showed that 74% of the people that went to Poupatempo units in 2008 did not consider using the pre-existing bureaucracy, an increase from 67% in 2006and 70% in 2007(Poupatempo and Ibope 2008).In sum, Poupatempo could be an example of how institutional bypasses can have a broader impact by eliminating dysfunctional institutions.In this aspect, however, it is too early to tell whether Poupatempo has been a success.
One way in which a bypass can change pre-existing institutions is by creating competition among public institutions, which may provide incentives for dysfunctional institutions to improve.Indeed, Poupatempo gave citizens an option to exit the pre-existing bureaucracy and go somewhere else, while at the same time keeping the traditional service provider in place.The increasing numbers of people attending Poupatempo was the main evidence that citizens were happy with it, but the traditional bureaucracy also had the option to become more attractive to these citizens by improving their service.This accorded with the reformers' hope that innovations produced by these reforms would have feedback effects, being later incorporated by the traditional bureaucracy, creating a virtuous cycle of reforms (interview with Daniel Annenberg, 2009).In some cases, this actually happened, such as the digitalization project at the identification card central office.
There was at least one case in which changes were drastic: the complete modification Finally, as to political economy problems, the bypass can reduce resistance among bureaucrats in the pre-existing institutions because their services, routines and obligations may not be majorly affected by the reforms in the short-term.It is acknowledged that bureaucrats might foresee that they will be losing power and influence in the long term, if the bypass succeeds, and they may resist its implementation for fear of being replaced by the new service.This lobby against the implementation of a bypass can turn into a boycotting scheme once the bypass is implemented.But even in this case, reformers are more able to implement innovations in the bypass than they could be in the pre-existing institution.In this regard, many (but not all) of the political economy problems are overcome or mitigated.

Notes
1 They often adopt Douglass North's definition of institutions: "Institutions are the rules of the game of a society, or, more formally, are the humanly devised constraints that structure human interactions" (North 1995, 23).
2 See Rezende (2009, note 3) for a full citation of these studies.
4 This number includes people that go to a unit and those who use the call centers.It does not include those who use the services via internet (e-Poupatempo).
5 Besides the experience of the centers of São Paulo and Bahia, this guide also considered the experiences of other states such as Amazonas (where the service was called Pronto Atendimento do Cidadão), Maranhão (Viva Cidadão), Rio de Janeiro (Rio Simples) and Rio Grande do Norte (Central do Cidadão).
6 The Unit of Strategic Management (Unidade de Gestão Estratégica) was subordinate to the State Secretary of Government and Strategic Management (Secretaria do Governo e Gestão Estratégica).It former was headed by Roberto Abunni and later by Dalmo Nogueira.For an explanation of this structure and the respective roles in the transformations implemented after 1995 in São Paulo, see Sanchez (2003).The adjunct secretary reports that there was an intense dialogue between the Unit and the Secretary (Interview with Dalmo Nogueira, 2009).

7
As the superintendent reported, some of the civil servants came from governmental departments, others did not.For instance, the first team had an IT person and a human resources person from the state-owned company for data processing of the State of São Paulo (Prodesp) and two architects from the state-owned company for metropolitan planning of the State of São Paulo (Emplasa).The manager and the person responsible for the information booklet, on the other hand, were in the private sector (Interview with Daniel Annenberg, 2009).
8 Sabesp is half state-owned, half a private company.The assessment of its efficiency in the context of this project does not refer to the delivery of water services itself, but exclusively to its ability to communicate to consumers efficiently.9 It is important to note that the identity card is the primary identification document for Brazilian citizens.
10 As the adjunct secretary of the State Secretary of Government and Strategic Management says, a great deal of these innovations were being discussed in Brazil since the 1970s within two institutions: Fundap and the Red Tape Ministry (Interview with Dalmo Nogueira, 2009).The focus of the state administration in 1995 was on the use of technology to improve governmental services.For a detailed description of this process, see Sanchez (2003).
11 The call centre works from Monday to Friday from 7am to 8pm, and Saturday from 6:30am to 3pm, which are longer hours of operation than most of the Poupatempo units.
12 As the superintendent reports, the salaries' increase varied according to the position that the civil servant exercised.The gratification was of R$ 500.00(U$$ 250.00) per month to the receptionists, R$ 700.00 (U$$ 350.00) to the tutors and R$ 900.00 (U$$ 450.00) to the supervisor (Interview with Daniel Annenberg 2009).It is important to note that the staff emphasizes that there were greater incentives in going to work on a pleasant place and in a new promising institution.Civil servants were proud of being part of Poupatempo (Interview with Dalmo Nogueira, 2009).
13 Information given by the Superintendência do Poupatempo, May 2009.
14 This is an estimate, as this data is not available.The estimate is based on the cost to build a brand new unit (R$ 6.5 millions), plus costs in the implementation of the program (building up the team, diagnosing problems in the offices of the pre-existing bureaucracies etc).
15 The superintendent indicated the importance of user's feedback (Interview with Daniel Annenberg, 2009).The secretary-adjunt indicated the governor's concern with external assessment (Interview with Dalmo Nogueira, 2009).
16 Political economy has been historically used to define a variety of different things.For the purposes of this paper, we will use the term as a synonym to public choice theory as applied to institutional reforms, which assumes that both reformers and those resisting reforms are all utility maximizing individuals, and shows how political forces (i.e.self-interested acts) affect the choice of policies, especially as to distributional conflicts.See Alesina (2007, 3) and Groenwegen (2008).
by making the provision of certain services exclusive of Poupatempo.

Mota Prado and Ana Carolina Chasin documents
and procedures to obtain services, so that citizens could gather the necessary documents in advance.Previously, citizens would be required to travel to a department just to obtain information about what they needed, returning later with the appropriate documents.Often times, there were misunderstandings or lack of accurate information, Two different communication networks were created for Poupatempo services: the Intragov and the Strategic System of Information.Created in 2001, the Intragov is a web-based network that uses data, voice and images and is available to public services in the entire state of São Paulo.Before that, the Poupatempo units had only internal communication systems (i.e. for different offices within each unit).The Strategic System of Information is also a web-based network that connects the governor's office with all the offices of the executive branch (Annenberg et al.Poupatempo also sought to improve the communication between government departments and citizens.For this purpose, it prepared booklets indicating all the required bpsr (2011) 5 (1) 19 11 -34 Mariana

innovative Was the Poupatempo Experience in Brazil? were
fully contracted out (i.e., the management of the entire unit is provided by a private company)(interview with Ilídio Machado, 2009).By 2009, around 45% of the employees of the Poupatempo (currently 6,695) were contract employees.13 mentions that contracting out process initially was used only for some services, but it expanded over time.Starting in 2006, some of the newly-created units (Osasco and Santos) bpsr (2011) 5 (1) 20 11 -34 How

Mariana Mota Prado and Ana Carolina Chasin de
(Noffs 2006ices of the Company for Urban and Housing Development in São Paulo (Companhia Desenvolvimento Habitacional e Urbano do Estado de São Paulo (CDHU)) to meet the standards of Poupatempo units.Before Poupatempo, there were no standardized services: they were executed in different ways in each department within the same company.Also, consumers would receive conflicting information if they called more than one department within the company.To join the Poupatempo project, CDHU had to restructure, unify, and standardize its services(Noffs 2006).This new system was eventually adopted within the entire company, not being restricted to the CDHU's offices at Poupatempo units.This case suggests that Poupatempo appears to have had some of the expected feedback effects on the traditional bureaucracy, promoting structural changes in the internal structures of the Brazilian State, but further research would be necessary to determine why this happened in this case, but not in other departments.As an institutional bypass, Poupatempo did not try to reform the pre-existing institutions.Its creators hoped that the success would have feedback effects and foster structural changes within the state's bureaucracy but it is not clear to what extent this has happened.The success of Poupatempo should not be reduced, however, to an evaluation of its capacity to eliminate dysfunctional institutions.Whether it has accomplished this or not, one cannot ignore the concrete benefits it has delivered to the population, as illustrated by the level of satisfaction of people who used it.Surveys conducted between 2001 and 2009 show that around 97% of the users support the project and 95% say that it provides Poupatempo.It argued that the project was an institutional bypass, which enabled it to overcome traditional obstacles to reforms.Because they are alternatives, not replacements,

How innovative Was the Poupatempo Experience in Brazil? institutional
bypasses can start as pilot projects and have reduced start-up costs.For instance, it requires only R$ 6.5 million to build and equip a new Poupatempo unit.As the Poupatempo project shows, reduced costs addresses the issue of scarce resources, as well as the risk of political resistance.Also, a bypass is equivalent to starting everything anew.This solves cultural barriers by putting bureaucrats and civil servants in a context devoid of the vices of the pre-existing bureaucracy.At the same time, a bypass is not as disruptive as replacing an entire system of bureaucratic offices.This non-disruptive rupture has three advantages.First, it allows for experimentation because the bypass does not tie the policymakers' hands to what is there already.Second, if successful, the bypass may provide useful feedback on potential improvements of the traditional system.Third, if not successful, reformers can easily abort the reform and go back to the existing system if something goes wrong (as like any other institutional reform, bypasses have no guarantee of success).In sum, it helps reformers deal with the unintended consequences of institutional change.
Thus, Poupatempo illustrates a type of institutional change or institutional reform not yet captured or properly described in the specialized literature, called institutional bypass.We hope this article is just the first step towards future research to investigate not only whether there are other examples of bypasses out there, but also the theoretical implications of this concept.interviews