)P%S 3c76 POLICY RESEARCH WORKING PAPER 3 070 Are You Satisfied? Citizen Feedback and Delivery of Urban Services Uwe Deicbmann Somik V. Lall The World Bank Development Research Group Infrastructure and Environment June 2003 I POLICY RESEARCH WORKING PAPER 3070 Abstract Citizen feedback is considered an effective means for expectations about service performance, influence a improving the performance of public utilities. But how household's satisfaction with service delivery. The well does such information reflect the actual quality of authors find that satisfaction increases with service delivery? Do so-called scorecards or report cards improvements in the household's own service status, a measure public service delivery accurately, or do finding that supports the use of scorecard initiatives. But personal and community characteristics have a significant the results also suggest that a household's satisfaction is impact on residents' assessment of service quality? influenced by how service quality compares with that of Deichmann and Lall investigate these questions using its neighbors or peers and by household level newly available household survey data on access to and characteristics such as welfare and tenure status. This satisfaction with selected public services in two Indian implies that responses in satisfaction surveys are at least cities-Bangalore and Jaipur. They develop a framework in part determined by factors that are unrelated to the where actual levels of services received, as well as service performance experienced by the household. This paper-a product of Infrastructure and Environment, Development Research Group-is part of a larger effort in the group to improve urban management through the collection, use, and public disclosure of spatially detailed information and analytic methods. The study was jointly funded by the World Bank and the U.K. DFID's Urban Knowledge Generation and Toolkits Program. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Yasmin D'Souza, room MC2-622, telephone 202-473-1449, fax 202-522-3230, email address ydsouza@worldbank.org. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at udeichmann@worldbank.org or slalll@worldbank.org. June 2003. (31 pages) The Policy Researh Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is toget the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the viewo of the World Bank, its Executive Directors, or the countries they represent. Prochiicerl hv Pqrrnerqhinq_ Canacirv Riiilding and Ourirench Are You Satisfied? Citizen Feedback and Delivery of Urban Services* Uwe Deichmann and Somik V. Lall Development Research Group, World Bank, Washington DC, USA This paper is part of a larger program to find ways to improve urban management through the collection, use and public disclosure of spatially detailed information and analytic methods, funded by the UEK DFID's Urban Knowledge Generation and Toolkits program and the World Bank Development Research Group. We thank Kenneth Chomitz, Michael Lokshin, Samuel Paul, and Zmarak Shalizi for helpful comments, and Ajay Suri and the team from the Society for Development Studies, Delhi, who implemented the household surveys. The authors can be contacted at Uwe Deichmann (udeichmann(aworldbank.org) or Somik V. Lall (slalll(aworldbank.or ), 1818 H Street NW, Washington, DC 20433, USA. Are you satisfied? Citizen feedback and delivery of urban services 1. Introduction Citizen feedback is often seen as an effective means for evaluating the quality of urban public service delivery. The rationale behind this approach is that collective or organized feedback can be used to demand accountability from providers of public services, especially when there are no alternate providers due to regulation and natural monopolies in the delivery of these services. One of the most widely cited examples of organized citizen feedback is the 'report card' or 'scorecard' where users are asked to rate their satisfaction with various aspects of service provision (Paul 1998). Examples include evaluation of service delivery in India, the Philippines, Ukraine, Malaysia, and the United States (Shah and Wagle 2001, World Bank 2001, UMP 2000). Report cards that solicit feedback can provide citizens an opportunity to shape the decisions that affect their lives. Stern (2002), among others, therefore suggests that such initiatives are a promising instrument for empowerment Since the objective of scorecards is to create public awareness as well as enhance responsiveness of service providers, it is useful to examine the extent to which such information reflects the actual quality of service delivery. Do these scorecards or report cards reflect accurately what citizens get in return for taxes or user charges, or are responses significantly influenced by personal and community characteristics? If responses are influenced by subjective factors, then it becomes difficult to compare service levels across user groups on the basis of feedback alone. For example, two individuals receiving the same service level may respond quite differently to questions about their satisfaction with service delivery, depending on their incomes or educational attainment. Similarly, relative disparities in service levels between individuals and their peer groups may influence satisfaction scores. At any given level of service, we may be less satisfied if we see that our neighbors receive better services. Such behavioral and psychological factors could distort evaluation of the actual service levels received and could therefore diminish the utility of citizen report cards in public policy making. It appears, however, that despite the increasing popularity of these approaches in the development community, we know little about the relative magnitude of objective versus subjective factors. Adding to the difficulties in evaluating the utility of scorecards is the frequent absence of clear benchmarks for public service provision. If there are no standards to which public service providers have committed, then evaluating whether or not the provider is doing a satisfactory job becomes a subjective process that is influenced by the idiosyncratic judgment of each service user. In this paper we investigate citizen satisfaction with various attributes of service delivery using newly available household survey data from two Indian cities. These surveys collected detailed information on households' satisfaction with various aspects of service delivery as well as information on actual service attributes. Our main objective is to examine the extent to which household and community characteristics influence a household's satisfaction with service delivery. Our basic premise is that satisfaction is determined by both the actual service levels and by expectations about service quality. The remainder of the paper is organized as follows: We present an overview of the existing literature on measuring satisfaction with service delivery in Section 2. In Section 3, we discuss some of the literature examining determinants of satisfaction. A model of satisfaction from service provision is presented in Section 4. We discuss the data in Section 5 and results in Section 6. Conclusions and policy implications are presented in Section 7. 2. Scorecards and benchmarking in public service provision Despite recent and sometimes controversial privatization initiatives, public ownership of utilities remains the most widely adopted model for the provision of basic services in developing countries (Roth 1987, Noll et al. 2000, Estache et al. 2001). For a number of reasons, services delivered by these institutions often remain inadequate. In many rural and urban areas, large proportions of the population have little or no access to public services. The quality of services for those who receive them is often poor -characterized by low quantities provided, inconveniences imposed on the users, frequent breakdowns and unresponsive providers. At the same time, few utilities are able to recover sufficient funds to cover the full cost of service provision, let alone to invest in improvement or 2 extension of services within their jurisdictions. Pricing is often politically motivated with little concern for cost of provision and distribution of benefits. Subsidies-intentionally or not-frequently tend to benefit better-off residents (Hentschel and Lanjouw 2000, Walker et al. 2000). What options do residents have to address these shortcomings in public sector performance? Hirschman's (1970) voice and exit are the main responses available to consumers confronted with inadequate goods and services. Exit is a viable strategy in competitive markets where alternative providers exist and the risk of market failure is limited. These conditions are rarely met in the basic services sectors, although private providers often play a limited role in compensating for inadequate public provision. For instance, private water vendors supply under-serviced areas in many developing country cities (e.g., Lovei and Whittington 1993) and own-provision of solid waste disposal services has been initiated in some urban areas (Dahiya 2003). However, many public services such as piped water supply, sewerage or electricity, require large capital investments to operate efficiently, have significant economies of scale and considerable network effects. Small-scale private or own provision therefore tends to be more expensive than the alternative of a well-regulated monopolistic supplier. Exit therefore is usually not a viable option for those who obtain public services and is, of course, an irrelevant issue for those who do not. Voice, in contrast, is an option both for households receiving inadequate service and for those without access. Generally, voice refers to the pressure that can be exerted on public service providers by well-organized citizen groups. This may occur either directly through complaint or protest, or indirectly by influencing or participating in the political process in cases where utilities are controlled by democratically elected public officials. During the last decade, citizen groups in a number of countries have championed the use of public feedback mechanisms to improve the performance of public sector entities. The rationale behind these efforts is that due to the monopolistic nature of public service provision and prevailing public apathy, public utilities typically lack the incentives to provide the highest possible service standards. Public disclosure of these shortcomings 3 will exert pressure on public service providers that can lead to improvements in their performance and increase the quality of life of their customers. A well documented example is the work by the Public Affairs Center (PAC) in Bangalore India (Paul 1998, 2002). The Bangalore report cards summarize citizens' assessment of services provided by public agencies or utilities such as the local development authority, water and sanitation board, transport authority and hospitals. In addition to overall satisfaction scores, the PAC scorecards solicit opinions about specific aspects of service provision. These include staff behavior, quality of service and communication of information. PAC also collects information on bribes paid in connection with service provision and household investments to cope with shortcomings in service provision such as water storage tanks or voltage stabilizers. Both add to the household's total cost of obtaining public services. PAC's work in Bangalore is reported to have triggered tangible efforts by the local government to improve service provision. It also led to similar work in several Indian cities (Balakrishnan and Sekhar 1998) as well as a country-wide initiative, the Millennial Survey of Public Services in India (PAC 2002). Scorecard initiatives have not been limited to developing countries. In industrial countries the desire to improve performance of local government functions has led to the establishment of so-called benchmarking initiatives (e.g., Samuels 1998). These assess performance internally but also attempt to measure the satisfaction of the users of public services. Other examples of citizen initiatives in both developing and western countries have been reviewed by the Institute of Development Studies (Goetz and Gaventa 2001). The accumulated evidence suggests that report cards, initiatives to amplify or coordinate citizen's voice, and benchmarking approaches can be an effective advocacy tool. There is limited experience, however, to conclude that citizen feedback initiatives are objective instruments for measuring local government performance. 3. Determinants of satisfaction Most scorecard initiatives seek to base their advocacy on empirical evidence mostly in the form of survey information that measures the satisfaction of citizens with various aspects of public service provision. Residents state whether or not they are satisfied or 4 they are asked to rank their satisfaction level on a predetermined scale. Subjective self- evaluation of perceived rankings, however, may not provide a completely accurate reflection of the reality faced by residents. The main constraints relate to measurement and context. The first issue is that reported satisfaction may not equal actual satisfaction which is unknown to anyone except the respondent. It would be virtually impossible to design a measurement approach that allows households to rank satisfaction unambiguously using'the same cardinal or ordinal scale. Any interpretation of such data, however, assumes that interpersonal comparisons of the benefits or utility derived from a given level of service are consistent. The problem is similar to the one faced when analyzing utility more generally. As reviewed by Ng (1997), many economists reject attempts to measure and compare utility. For this reason, studies that analyze subjective variables such as satisfaction or happiness scores have traditionally been met with skepticism among economists. This has changed recently, in part due to the evidence accumulated by psychologists on the validity of subjective responses on topics related to utility, welfare or satisfaction (Diener et al. 1999). The second main caveat concerning the interpretation of scorecard responses is that reported satisfaction may be influenced by a multitude of contextual factors only some of which will be related to the characteristics of the service itself. Other factors 'such as personal and community characteristics will also influence perceptions about the quality of the service received. There has been little empirical analysis of the determinants of satisfaction with public services. But there exist two strands of literature that are of relevance to our questions: studies on the determinants of happiness in general and satisfaction with economic status in particular, and the marketing literature on customer satisfaction. Concerning the former, Blanchflower and Oswald (2003) provide a concise list of the main determinants of self-reported well-being: circumstances, aspirations, comparison with others, and what they refer to as a person's "baseline happiness". In the context of public service delivery, the first two of these factors correspond to performance and expectations. Performance is the actual,' quantifiable level of service received by the household. If scorecard initiatives are to be credible, performance should be the main determinant of perceived service delivery and thus satisfaction. Perceptions, however, are influenced by expectations-the benchmark against which a person measures actual experiences. Furthermore, 5 expectations may be formed by personal characteristics such as education, but they may also be influenced by the household's comparison of own service access with that of others. "Baseline satisfaction" is best considered as an idiosyncratic component that describes the willingness to put up with adverse conditions, or conversely, the tendency to complain. We could term this the squeaky wheel factor. Given the close relationship between determinants of happiness and satisfaction, research on happiness and welfare clearly has much to teach us in understanding satisfaction with basic service provision. The study of perceived welfare or happiness originated in psychology but has subsequently also attracted the interest of economists (see Veenhoven 1996, Dixon 1997, Diener et al. 1999, Kahneman and Tversky 2000, Lokshin and Ravallion 2001, Blanchflower and Oswald 2003). One of the first and most significant findings was Easterlin's (1974) proposition that increases in income do not considerably increase a person's happiness. Oswald (1997), for example, confirms these findings in a study of happiness over time in the US and Britain. While average income has increased considerably over the last several decades, the proportion of survey respondents who stated that they are happy has increased much less. Thus, the general conclusion is that "money does not buy happiness, or at least not much" (Ng 1997). Instead, happiness in those countries was found to be correlated with being married, employed and seems to be U-shaped in age with a minimum in the early thirties. A common theme in this literature is the importance of reference group comparison. Easterlin (1974) proposes that relative income is more important than absolute income in determining happiness. In the context of job and income satisfaction, Clark and Oswald (1996) analyze satisfaction with income among British workers. They find an inverse relationship between workers' income satisfaction and their reference or comparison income, which is derived as the predicted income for the worker based on a standard earnings equation using the entire cross-sectional data set. Workers' satisfaction, according to this analysis, is less influenced by absolute income levels, than by how that income compares to that of similar persons with similar jobs. This confirmed earlier work on reference group effects that indicates that unemployed workers in Britain who live in areas of high unemployment show less mental distress-i.e., are less unhappy-than 6 those in areas of low unemployment (Clark and Oswald 1994). Frank (1997) discusses the issue of reference groups more generally and concludes that satisfaction is greatly influenced by the individual's or household's relative position. One manifestation of this is the desire of people to "keep up with the Jones n. Another strand of the satisfaction literature has its origin in the field of consumer research and marketing. Customer satisfaction is a pre-requisite for customer loyalty. Consequently, there has been a lot of interest in determining what drives consumer satisfaction. This research draws extensively on the psychology literature. Marketing research has identified two main models of satisfaction that differ significantly in how performance and expectations are linked (Johnson et al. 1995 and 1996; see also Sasser et al. 1978, Boulding et al. 1993, Driver 2002). The disconfirmation model assumes that satisfaction is the difference between performance and expectations. If performance is equal to or higher than expectations, the customer is satisfied. If expectations exceed performance, the customer is dissatisfied. This implies that the absolute level of product or service performance does not need to influence satisfaction directly. For instance, a poor household with low expectations may find that relatively low service quality still exceeds these expectations. The household may thus be as satisfied as a rich household with high expectations and much better service availability. Expectations thus serve as a standard of reference that reflects opinions about what the service provider should deliver. Obviously, these could be far from realistic. An alternative approach that has gained increasing following in marketing is the performance model in which both product or service quality and expectations affect satisfaction levels positively. The performance model assumes that expectations serve as an "anchor" to the evaluation of performance. Performance, as perceived by the customer then determines satisfaction levels. Customers are assumed to continuously adjust their expectations as they experience the products or services. They will thus have a (more or less) realistic idea of what the service provider will deliver. Expectations and performance will therefore never vary by much and both could therefore be positively related to satisfaction levels. Johnson et al. (1996) point out that this model is similar to 7 the cognitive process of anchoring and adjustment described by Tversky and Kahneman (1974).' 4. A model of satisfaction from service provision The evidence reviewed in the previous section informs our model of satisfaction with basic urban services, in this case water supply. In this model, utility-i.e., satisfaction with service delivery-is a function of actual service performance and the person's expectations about service performance. Thus in our framework, U =f(A, E), where U represents utility or the perceived satisfaction derived from provision of a service. A represents a vector of characteristics describing the quality and quantity of the service received by the individual, and E is a vector of individual and community characteristics that determine a person's expectations about service performance. Service performance is the actual level of service provided and available to the household. It consists of measurable aspects such as the quantity provided, the frequency at which the service is available and the quality of the product delivered. Most of these can be measured with some degree of accuracy to yield indicators such as the quantity of water provided to households, the number of days on which water is available or the purity of the water coming out of the tap. But public service delivery also has aspects that are less easily quantified, for example, the responsiveness of the provider in addressing service problems or the courtesy of its representatives in their interaction with the general public.2 In evaluating satisfaction with public services, individuals compare actual levels of service performance with their expectations about service performance. There are two ways through which these expectations might be formed. The household may compare the actual level of service received to what the public utility will likely provide given past experience. For instance, poor performance in the past will lead the household to expect XSee also Kahneman et al. 1999. Johnson et al. (1996) discuss two additional models. The rational expectations model, assumes-along the lines of the original rational expectations theory (Muth 1961)- that perceived performance and expectations are essentially identical. For complex products or services in which performance influences satisfaction directly, they argue that performance and expectations have a positive relationship without expectations affecting satisfaction directly. These issues are extensively studied in the marketing literature on the provision of private goods (Driver 2002). 8 poor performance in the future. Alternatively, the expectation may be formed by some assessment of what the public utility should provide according to some idealized, but possibly unrealistic, benchmark. This benchmark might be determined arbitrarily, such as around-the-clock water supply in a city where households typically receive water for a few hours every other day. In the context of public service provision, we initially favor the performance model of consumer satisfaction. We believe that households who experience actual levels of service delivery every day are likely to anchor their expectations to the actual performance of the utility. They are thus not assumed to have unrealistic expectations about what providers could potentially deliver. Direct measurement of expectations is clearly as difficult as measuring satisfaction itself. We assume that expectations are influenced by various individual and/or household attributes. These may include age (Duffy 2000), education (Clark and Oswald 1996), and ethnicity (Bartel et al. 1996), as well as welfare status, tenure arrangements and gender. Education, for instance, may increase aspirations and consequently expectations concerning the level of services received in return for user fees or taxes. Furthermore, more educated citizens might be less reluctant to confront providers of public services and more willing to complain about inadequate service provision. Higher education levels might therefore be associated with lower satisfaction scores. Gender may influence expectations in two possibly contradictory, ways. If there are intra-household differences in the coping costs of inadeqte water supply-for example where households need to rely on public taps-it is often women who bear the burden of obtaining water several times a day. Their perception of services ieceived will be different and their dissatisfaction will consequently be larger than that for household members who do not share this burden. For the same reason, however, women may be better informed about the actual level of service provision. If the performance model holds, they should be more likely to anchor their expectations around actual service delivery. In that case, women may not show lower satisfaction levels, all else being equal. We also assume that home owners have higher expectations concerning public service provision than short-term renters, since the quality of services directly affects the value of the dwelling unit and because owners are likely to remain in the same dwelling for a longer period of time. 9 We define satisfaction or utility from water supply as a standard microeconomic model of household behavior: U= u(p(w), i, j, x-c(w)) (la) where p(w) represents the attributes of service performance such as the quantity and quality of water provision andj represents additional supply characteristics, such as the type and location of water supply. Further, i represents individual and household specific characteristics influencing expectations about water supply provision and x-c(w) is the utility from consumption of other goods less the expenditure on water. In this model, satisfaction or perceived utility from water supply is increasing in service availability and consumption of other goods, subject to expenditures on water. This is because utility will decrease with higher cost of water procurement, c(w). Relative service performance in reference group While we believe that households have realistic expectations concerning service delivery, evidence from the psychology and happiness literature suggests that additional factors influence expectations and thus satisfaction levels. Moving beyond the traditional utility maximizing framework in (l a), we assume that relative performance-the quality of service received in comparison with that received by others-will modify expectations and therefore perceived performance and thus satisfaction. We modify (la) to consider service performance in the household's reference group: U = u(p(w), p(w*), i, j, x-c(w)) (Ib) where p(w *) represents relative service performance in the reference group. Perceived utility from water supply is an increasing function of a household's own service performance and is also influenced by reference group performance. For instance, utility could be an increasing function of the difference or ratio between service performance for the household and that of its reference group. Thus, if a household receives poorer service than its peer group, it is likely that its expectation is higher than actual performance and satisfaction is therefore low. Satisfaction is likely to be higher, if the household's peer group receives equal or worse services, even if the absolute level of service provision is 10 inadequate. In other words, shortcomings in service delivery may be more acceptable when all neighbors or social acquaintances are in the same situation, but unacceptable when the household is the only one suffering from low quality services. This suggests that relative deprivation, in addition to the absolute service level, is critically important in determining satisfaction. Incorporating reference group service levels thus has the effect of moving from the performance model (assuming "will expectations") towards the disconfinnation model ("should expectations"). Figure 1 illustrates the assumptions underlying our model of reference group effects on satisfaction. Service levels in the household's reference group are on the x-axis and own service levels are on the y-axis. Utility, or in this case satisfaction from services increases from the south east to the north west section of the graph. This means that satisfaction increases as own service levels improve in comparison to the reference group (indifference curve ICO to IC2). Point A on ICO shows satisfaction levels (or utility) when the household receives the same service level as its reference group (z). At A, performance equals expectations, and the household will be satisfied by the service level received. Satisfaction decreases as we move to indifference curve IC 1 when service levels-in the reference group (z+q') are higher than that for the household (z-q, where q does not need to equal q'). At point B, the "should expectations" are higher than performance and the household perceives relative deprivation, thereby reducing its satisfaction from service provision. On the other hand, at point C on curve IC2, service levels for the household (z+q) are higher than the reference group (z-q'). In this case, performance is better than expectations and the household perceives higher satisfaction from service provision. Figure 1: Own versus reference group service performance IC2 'Co z+q --- 0 z --q -- 0 z-q ------- O I I\ I t I I I I I I I 0 z-q z z+q' reference group service performance Specification As described in equation (lb) and Figure 1, satisfaction from services is a function of a household's own service levels and the relative service level in the reference group controlling for other individual or group characteristics. Our interest is in explicitly measuring relative performance-i.e., reference group performance relative to own performance. Rather than using reference group performance (P) on its own, as for example in the case of comparison income in Clark and Oswald (1996), we therefore use the 'ratio comparisons' model specification suggested by Clark and Oswald (1998). Satisfaction from services is thus defined as: s; ='Pi + T'(-5,)+ v7+Y1 + g (2) where Si is the household's satisfaction or utility from attributes of service delivery, Pi measures the household's own service performance and P* represents service levels in the household's reference group. The term (piI) is the ratio of the service level between the household and its reference group. The household receives satisfaction from 12 improvements in its own service levels (a') as well as from surpassing its peers (') when P,>P . Conversely, satisfaction is reduced if Pi

LR 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Robust standard errors in parentheses + significant at 10 percent; * significant at 5 percent; ** significant at I percent 29 Table 5:Satisfaction estimation for duration of water supply (Jaipur) Reference group defined geographically Reference group defined by language/welfare Household head Respondent Household head Respondent charcteristics characteristics characteristics characteristics Hours per day 0.096 0.095 0.097 0.094 0.040 0.030 0.041 0.030 - (0.007)i* (0.008)** (0.007)** (0.008)** (0.01 1)** (0.01 1)** (0.010)** (0.0 )t** Relative HPD -0.078 -0.068 -0.080 -0.069 0.387 0.429 0.385 0.428 (0.053) (0.044) (0.053) (0.044) (0.072)** (0.071)** (0.072)+* (0.071)** .Welfare status 0.450 0.427 0.405 0.375 0.383 0.353 0.342 0.305 (0.066)*0 (0.068)** (0.064)** (0.067)** (0.067)** (0.070)+* (0.066)** (0.068)** Tenure status -0.168 -0.169 -0.196 -0.197 -0.166 -0.164 -0.195 -0.192 (0.103) (0.102)+ (0.102)+ (0.101)+ (0.104) (0.104) (0.103)+ (0.103)+ Household size -0.267 -0.262 -0.273 -0.269 -0.239 -0.236 -0.246 -0.244 (0.072)*$ (0.073)** (0.071)** (0.071)** (0.073)** (0.073)** (0.071)** (0.071)** Education -0.049 -0.047 0.028 0.037 -0.046 -0.042 0.020 0.032 (0.067) (0.068) (0.070) (0.070) (0.068) (0.068) (0.070) (0.070) Age -0.196 -0.213 -0.065 -0.070 -0.194 -0.212 -0.064 -0.070 (0.114)+ (0.113)+ (0.074) (0.075) (0.115)+ (0.115)+ (0.074) (0.075) Sex -0.136 -0.121 0.059 0.062 -0.143 -0.118 0.048 0.056 (0.118) (0.120) (0.060) (0.060) (0.119) (0.121) (0.060) (0.060) Source type I 0.332 0.339 0.494 0.499 (0.1I99)+ (0.1I99)+ (0. 196)* (0.196)* Source type 2 -0.033 -0.064 -0.123 -0.152 (0.144) (0.144) (0.147) (0.146) Source type 3 -0.298 -0.288 -0.162 -0.155 (0.222) (0.223) (0.222) (0.222) Log Likelihood -1953.444 -1947.393 -1956.695 -1950.561 -1939.60 -1930.480 -1943.112 -1933.684 Likelihood ratio 579.017 591.120 577.646 589.914 606.692 624.944 604.811 623.668 Prob > LR 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Robust standard errors in parentheses + significant at 10 percent; * significant at 5 percent; *4 significant at 1 percent 30 Table 6: Marginal effect on the probability of being satisfied with the duration of water supply (Bangalore) (1) (2) (3) (4) Hours per day 0.008 0.013 -0.003 0.000 Relative HPD 0.064 0.055 0.177 0.174 Welfare status 0.157 0.094 0.160 0.096 Tenure status -0.066 -0.059 -0.065 -0.058 Household Size -0.115 -0.081 -0.121 -0.085 Education 0.026 0.009 0.025 0.007 Age 0.060 0.025 0.060 0.025 Sex 0.104 0.106 0.097 0.100 Fixed Effects No Yes No Yes Reference Group Geographic Geographic Socio- Socio- economic economic Table 7: Marginal effect on the probability of being satisfied with the duration of water supply (Jaipur)' (1) (2) (3) (4) Hoursperday 0.038 0.037 .0.016 0.012 Relative HPD -0.031 -0.027 0.151 0.167 Welfare status 0.176 0.167 0.150 0.138 Tenure status -0.066 -0.067 -0.065 -0.065 Household Size -0.105 -0.103 -0.093 -0.092 Education -0.019 -0.018 -0.018 -0.017 Age -0.077 -0.084 -0.076 -0.083 Sex -0.053 -0.047 0.055 -0.045 Fixed Effects No Yes No Yes Reference Group Geographic Geographic Socio- Socio- economic economic Using individual level characteristics of the household head. 31 I Policy Research Working Paper Series Contact Title Author Date for paper WPS3043 The Incentive-Compatible Design Thorsten Beck May 2003 K. Labrie of Deposit Insurance and Bank 31001 Failure Resolution: Concepts and- Country Studies WPS3044 Impregnated Nets Cannot Fully Mead over May 2003 H. Sladovich Substitute for DDT: Field Bernard Bakote'e 37698 Effectiveness of Malaria Prevention Raman Velayudhan in Solomon Islands Peter Wilikai Patricia M. Graves WPS3045 Causes and Consequences of Civil Klaus Deininger May 2003 M. Fernandez Strife: Micro-Level Evidence from 33766 Uganda WPS3046 Migration, Spillovers, and Trade Ross Levine May 2003 E. Khine Diversion: The Impact of Sergio L. Schmukler 37471 Internationalization on Stock Market Liquidity WPS3047 Comparing Mortgage Credit Risk Robert Buckley May 2003 0. Himid Policies: An Options-Based Approach Gulmira Karaguishiyeva 80225 Robert Van Order Laura Vecvagare WPS3048 Targeted Transfers in Poor Countries: Martin Ravallion May 2003 P. Sader Revisiting the Tradeoffs and Policy 33902 Options WPS3049 Hidden Impact? Ex-Post Evaluation Shaohua Chen May 2003 P. Sader of an Anti-Poverty Program Martin Ravallion 33902 WPS3050 The Changing Financial Landscape: Wafik Grais May 2003 R. Vo Opportunities and Challenges for Zeynep Kantur 33722 the Middle East and North Africa WPS3051 Regional, Multilateral, and Unilateral Glenn W. Harrison May 2003 P. Flewitt Trade Policies of MERCOSUR for Thomas F. Rutherford 32724 Growth and Poverty Reduction in David G. Tarr Brazil Angelo Gurgel WPS3052 Long-Run Impacts of China's WTO Kym Anderson May 2003 S. Lipscomb Accession on Farm-Nonfarm Income Jikun Huang 87266 Inequality and Rural Poverty Elena lanchovichina WPS3053 Economic Impacts of China's Elena lanchovichina May 2003 S. Lipscomb Accession to the World Trade William Martin 87266 Organization WPS3054 The New Comparative Economics Simeon Djankov May 2003 G. Sorensen Edward Glaeser 37088 Rafael La Porta Florencio Lopez-de-Silanes Andrei Shleifer WPS3055 The Global Growth of Mutual Funds Deepthi Fernando May 2003 A. Yaptenco Leora Klapper 31823 Viktor Sulla Dimitri Vittas Policy Research Working Paper Series Contact Title Author Date for paper WPS3056 Avoiding the Pitfalls in Taxing Patrick Honohan May 2003 A. Yaptenco Financial Intermediation 31823 WPS3057 Teaching Adults to Read Better and Helen Abadzi May 2003 H. Abadzi Faster: Results from an Experiment 80375 in Burkina Faso WPS3058 Working for God? Evaluating Service Ritva Reinikka May 2003 H. Sladovich Delivery of Religious Not-for-Profit Jakob Svensson 37698 Health Care Providers in Uganda WPS3059 Do Capital Flows Respond to Risk Cesar Calder6n May 2003 T. Tourougui and Return? Norman Loayza 87431 Luis Serven WPS3060 World Market Integration through the Rui Alb,querque May 2003 T. Tourougui Lens of Foreign Direct Investors Norman Loayza 87431 Luis Serven WPS3061 Reciprocity in Free Trade Caroline Freund May 2003 P. Flewitt Agreements 1 32724 WPS3062 Labor Effects of Adult Mortality in Kathleen Beegle May 2003 E. de Castro Tanzanian Households 89121 WPS3063 Trade Liberalization, Firm Paolo Epifani May 2003 P. Flewitt Performance, and Labor Market 32724 Outcomes in the Developing World: What Can We Learn from Micro-Level Data? WPS3064 Trade Policy, Trade Volumes, and Ana M. Fernandes May 2003 A. Fernandes Plant-Level Productivity in Colombian 33983 Manufacturing Industries WPS3065 Ghost Doctors: Absenteeism in Nazmul Chaudhury May 2003 H. Sladovich Bangladeshi Health Facilities Jeffrey S. Hammer 37698 WPS3066 The Potential Demand for and Shreelata Rao Seshadri May 2003 H. Sladovich Strategic Use of an HIV-1 Vaccine P. Subramaniyam 37698 in Southern India Prabhat Jha WPS3067 The Mini-Integrated Macroeconomic Pierre-Richard Agenor May 2003 M. Gosiengfiao Model for Poverty Analysis: A Framework 33363 for Analyzing the Unemployment and Poverty Effects of Fiscal and Labor Market Reforms WPS3068 Poverty and Economic Growth in Heba El-laithy June 2003 P. Sader Egypt, 1995-2000 Michael Lokshin 33902 Arup Banerji WPS3069 International Migration, Remittances, Richard H. Adams, Jr. June 2003 N. Obias and the Brain Drain: A Study of 24 31986 Labor-Exporting Countries