WPS8673


Policy Research Working Paper                    8673




Political Selection and Bureaucratic Productivity
                             James Habyarimana
                               Stuti Khemani
                                 Thiago Scot




Development Economics
Development Research Group
December 2018
Policy Research Working Paper 8673


  Abstract
 Economic theory of public bureaucracies as complex orga-                           data in an institutional context of district governments in
 nizations predicts that bureaucratic productivity can be                           Uganda, which is typical of the local state in poor countries.
 shaped by the selection of different types of agents, beyond                       The paper measures traits such as the integrity, altruism,
 their incentives. This theory applies to the institutions of                       personality, and public service motivation of bureaucrats
 local government in the developing world, where nation-                            and politicians. It finds robust evidence that higher integrity
 ally appointed bureaucrats and locally elected politicians                         among locally elected politicians is associated with substan-
 together manage the implementation of public policies                              tively better delivery of public health services by district
 and the delivery of services. Yet, there is no evidence on                         bureaucracies. Together with the theory, this evidence sug-
 whether (which) selection traits of these bureaucrats and                          gests that policy makers seeking to build local state capacity
 politicians matter for the productivity of local bureaucracies.                    in poor countries should take political selection seriously.
 This paper addresses the empirical gap by gathering rich




 This paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the
 World Bank to provide open access to its research and make a contribution to development policy discussions around the
 world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors
 may be contacted at skhemani@worldbank.org.




         The Policy Research 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 to get 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 views of the International Bank for Reconstruction and Development/World Bank and
         its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.


                                                       Produced by the Research Support Team
   Political Selection and Bureaucratic Productivity∗
            James Habyarimana                    Stuti Khemani               Thiago Scot




   ∗
     We are indebted to Punam Chuhan-Pole and Andrew Dabalen for inviting this work to generate policy-
relevant research on governance; to Christina Malmberg-Calvo and Barbara Magezi Ndamira of the World
Bank country team for welcoming us to Uganda to undertake it; and, to Fred Finan for providing us with
survey modules on selection traits and advising this work. We thank Samuel Berlinski, Tito Cordella, Dan
Roggers, Luis Serven, Razvan Vlaicu, Alwyn Young, and seminar participants at the IDB and the World
Bank 2018 ABCDE Conference for valuable comments. We thank Jessica Bayern for excellent research as-
sistance in implementing the survey, Amshika Amar for preliminary analysis, and Ken Omondi and Swati
Raychaudhuri for exceptional administrative assistance. This work was made possible through nancial sup-
port from the Knowledge for Change Program, the Strategic Research Program, and the Research Support
Budget, for which we are very grateful. The views expressed here are those of the authors and not neces-
sarily of the World Bank, its Executive Directors, or the governments they represent. Stuti Khemani (corre-
sponding author): skhemani@worldbank.org. Thiago Scot, thiago_scot@berkeley.edu. James Habyarimana,
jph35@georgetown.edu
1        Introduction

The impact of anti-poverty and pro-growth policies and spending programs depends upon
how well they are implemented or delivered by government bureaucracies. The quintessen-
tial “delivery unit” in developing countries consists of local governments helmed by ap-
pointed bureaucrats and locally elected politicians. Institutional arrangements in the local
state, or the last mile at which delivery and implementation happens across countries such
as India, Indonesia, Nigeria, Brazil, Tanzania and Uganda, provide space for locally elected
politicians to facilitate and monitor the administration of nationally funded programs (Ah-
mad and Brosio, 2006 and 2009). These local institutions correspond to the setting exam-
ined in principal-agent theory of the public sector: multiple principals and agents; and,
complexity of tasks and objectives (Tirole, 1994; Dewatripont, Jewett and Tirole, 1999;
Dixit, 2002). The theory suggests that productivity in such organizations can be enhanced
through organizational design that increases motivation and improves selection of agents,
going beyond incentives (Besley and Ghatak, 2005; Acemoglu, Kremer and Mian, 2007;
Akerlof, 2017). However, there is little to no empirical evidence on whether selection of
agents matters, and furthermore, which traits matter for performance.1 We provide the
 rst evidence on how political and bureaucratic selection traits correlate with the produc-
tivity of public sector organizations. We measure traits of local politicians and bureaucrats
in an institutional context which is typical of the local state in poor countries–district gov-
ernments in Uganda–and examine whether these measures are robustly and substantively
correlated with variation in service delivery by these districts.
        Surveys were undertaken during September-December 2015 in 75 of Uganda’s 112 dis-
tricts at that time,2 the relevant local jurisdiction where both directly elected district politi-
cians and nationally appointed bureaucrats share responsibility for implementing public
    1
      Finan, Olken and Pande (2015) review a growing literature which measures selection traits of agents, fo-
cusing on frontline service providers, such as health workers, and their individual productivity. For example,
Donato et al (2017) and Callen et al (2015) nd that public health workers with better personality traits are
less likely to be absent from their jobs. There is no evidence yet available in the literature on whether selec-
tion traits of bureaucrats and politicians who manage public sector organizations matter for the productivity
of those organizations.
    2
      The number of districts in Uganda has been changing, due to the splitting of districts to create more new
ones (Green, 2010; Grossman and Lewis, 2014). The current number is closer to 121.


                                                       1
policies and delivering services. Modules available from the literature to measure integrity,
altruism, cognition, risk-aversion, personality traits, and public service motivation, were
administered to 1,357 district bureaucrats and 770 district politicians. Administrative data
available on the implementation of nationally mandated public health programs serve as a
measure of productivity of the district organization. These data reveal that the integrity of
politicians is a robust predictor of district productivity in delivering health services. The
size of the correlation is substantively large: a 1 standard deviation higher average integrity
among local politicians is associated with a 0.2 to 0.4 standard deviation higher measure of
performance in health service delivery. In terms of concrete indicators of health coverage,
an increase of 1 s.d. in the average integrity of politicians is correlated with a 4.4 percent-
age point increase in child delivery at government facilities (9% increase at the mean); a
4.7 p.p. increase in share of households with latrines (6.2% at the mean); a 3.8 p.p. increase
in share of pregnant women preventively treated for malaria (6.7% at the mean) and a 2.8
p.p. increase in pregnant women having at least 4 antenatal care visits (8.2% at mean).
       In contrast to the correlation with political quality, we nd no signi cant correlation of
service delivery with measures of electoral competition such as the concentration of vote
shares of competing candidates and the margin of victory. At the same time, we nd that
political integrity tends to be higher in places where there is greater electoral competition.
This is suggestive evidence that competition can play a role through improving selection
(such as, by enabling the election of politicians of higher integrity) rather than through in-
centives alone. Our results are consistent with other research that incentives generated by
electoral competition may play an ambiguous role in public service delivery.3 Our evidence
also brings into the fore a hitherto neglected reason why quality (selection) of politicians
might matter beyond incentives – because public sector bureaucracies are complex orga-
   3
     Other research has cast doubt on the role of electoral incentives in improving outcomes. For example,
political incentives to seek re-election can distort the allocation of public funds from what a benevolent social
planner would choose (Finan and Mazzocco, 2017). Political parties can respond to greater electoral pressures
by elding criminals as candidates (Aidt, Golden and Tiwari, 2015), or by increasing vote-buying (Cruz,
Keefer and Labonne, 2016; Khemani, 2015), rather than by improving services. Afridi et al (2017) provide
both theory and consistent evidence that corruption in developing countries can increase with electoral
competition. Consistent with our results, Grossman and Michelitch (2018) nd that Ugandan local politicians
do not respond to increased incentives (through a transparency intervention) in competitive constituencies
by improving service delivery.


                                                       2
nizations whose performance depends upon the interaction between a large number of
principals and agents. Stronger electoral incentives of one set of actors – politicians – may
not generate the actions that are conducive to organizational or team productivity.
       The evidence we contribute on political integrity lends broad support to the theory of
public bureaucracies as complex organizations. In district governments in Uganda, as in
local government agencies of most poor countries, two types of agents share responsibility
for managing a swathe of public spending programs and implementing policies–nationally
appointed bureaucrats and locally elected politicians.4 While the nationally-appointed bu-
reaucrats are expected to bring the technical competence needed for public sector man-
agement, the locally elected politicians are supposed to play a role in monitoring and fa-
cilitating service delivery, especially by liaising with citizens and communities, increasing
public awareness and winning public support. The cadre of politicians whose quality (in-
tegrity) is revealed by our data as a signi cant correlate of service delivery–locally elected
district councilors in Uganda—are elected from constituencies (sub-counties) at the front-
lines of service delivery, where health clinics operate and public health workers perform
their tasks. Monitoring these health service delivery units is cited as one of the signi -
cant duties that local councilors are expected to perform (Grossman and Michelitch, 2018).
However, councilors have been described as typically failing to perform their duties of po-
litical oversight (Ra er, 2016). Our nding thus suggests that where these local politicians
are of better quality (higher integrity), they undertake their monitoring and facilitating
roles more e ectively, thereby contributing to improvements in service delivery. While
prior work has focused on the policy-making role of more powerful political leaders, at
national or regional levels, our work shows that the quality of more local politicians, who
wield facilitating and monitoring powers over service delivery, also matters.
       This evidence on political integrity is consistent with theory on the importance of po-
litical selection. The paper thus contributes to lling a large gap between economic theory
and empirical evidence on political selection (Besley, 2006 and Dal Bo et al, 2017, provide
   4
     Section 2 of the paper reviews the institutional arrangements in Uganda, and shows how it is typical of
local service delivery arrangements in developing countries (Ahmad and Brosio, 2006 and 2009).



                                                     3
reviews). The identity and characteristics of leaders can matter because the political institu-
tions through which leaders gain o ce (such as elections) o er only incomplete contracts,
and lack of enforceability by third parties (Osborne and Slivinski, 1996; Acemoglu, 2003).
Furthermore, the quality of politicians can matter because, as leaders, they can play a role
in shaping social norms and policy preferences (Acemoglu et al, 2015; Bidner and Francois,
2013). Yet, despite strong theoretical reasons to care about political selection, there is little
evidence of whether and how it matters. The few empirical studies available on political
selection are constrained by not having access to data on typically unobservable dimen-
sions of quality, such as integrity or honesty, and altruism or a sense of civic duty, which
may not be captured by available measures such as education (Ferraz and Finan, 2009).5
Indeed, in our own data we nd only weak and small correlations between education and
di cult-to-observe personality and behavioral traits.6 Theory suggests that self-selection
of political candidates on these di cult-to-observe variables, speci cally, integrity and al-
truism, can play an important role in governance (Bernheim and Kartik, 2014). We nd, in
fact, that political integrity matters for service delivery.
       We recognize and emphasize throughout the paper that this study can only report ro-
bust correlations, not a causal relationship. We cannot identify an exogenous institution
that drives variation in political selection. It may be that some omitted variable, such as
citizen preferences for public goods, is correlated with better quality citizens entering the
political market as candidates, better quality candidates getting elected, and with better
service delivery by health workers. We argue that the way to interpret the correlation we
 nd in the data is arising from the interaction between citizens, local politicians and ser-
vice providers, as in the framework provided in World Bank (2016a) of inter-dependent
principal-agent relationships in government. This interaction yields an equilibrium in
which citizen demand, political selection, and behavior of health workers are all endoge-
nously determined. The correlation between political integrity and local state productiv-
   5
      One strand of the empirical literature on political selection has focused on the phenomenon of adverse
selection of criminals as political candidates in India (Aidt, Golden and Tiwari, 2015; Asher and Novosad,
2018; George, Gupta and Neggers, 2018).
    6
      Correlations reported in Table 3.



                                                     4
ity is thus driven by variation across the 75 di erent equilibria of district governments.
We argue and provide supporting evidence to interpret the correlation as follows: in dis-
tricts where higher integrity politicians are endogenously selected, these politicians exert
greater e ort towards public goods than rent-extraction; as a result, the local implementa-
tion of national health programs improves. Even if omitted variables that capture citizens’
political preferences or prevailing social norms are the underlying driver of variation in
political selection, the proximate channel of in uence of political integrity on service de-
livery is through the e ort exerted by selected politicians. We show that the correlation
with political integrity holds even after accounting for plausible sources of variation in
voter demands (such as, partisan attachment to national leaders, and the presence of local
radio), thus suggesting that local politicians have a direct or proximate channel of in u-
ence on service delivery, apart from other channels through which citizens might in uence
service delivery without going through politicians.
   We further provide qualitative evidence and support from other studies in Uganda of
the role of local councilors in monitoring local health facilities, and how they can play
these roles in obstructionist, rent-seeking ways (such as shielding absent health workers
from disciplinary action, or pilfering funds at the health clinic) versus facilitating better de-
livery (such as encouraging health workers to show-up, and monitoring that clinic funds
are appropriately used). We suggest that the measure of integrity is capturing the procliv-
ity of politicians to exert e ort in one type of action compared to another–whether in the
extraction of private rents versus the promotion of broader service delivery. Higher in-
tegrity politicians are more motivated to pursue public goods and therefore more e ective
in improving public health service delivery.
   Nevertheless, omitted variables that explain both integrity and the ability of politicians
to in uence service delivery could be biasing the coe cient on integrity that we estimate.
Random assignment of higher integrity politicians to districts may yield di erent estimates
of impact on service delivery. However, random assignment of integrity may not be the
appropriate way to test theories where political selection arises endogenously. Learning
from correlations revealed by the data is valuable to test theories of endogenous relation-

                                               5
ships that arise in equilibrium, and to inform further theoretical development. For exam-
ple, models of political selection have been quite separate from models of agent selection
in complex organizations of the bureaucracy. Our results suggest scope for further de-
veloping a theory of how these two di erent types of agents, bureaucrats and politicians,
interact in government agencies. Our study provides unique new evidence to understand
the relative characteristics of these agents.
   We nd systematic di erences across several dimensions of quality between politicians
and bureaucrats, and, in which features of quality of these two types of agents matter for
service delivery. For example, bureaucrats are more educated than politicians and educa-
tion of bureaucrats is correlated with service delivery; in contrast, education of politicians
is not correlated with service delivery. These di erences in education between bureau-
crats and politicians may be viewed as simply con rming what one might have suspected
given the local institutional context where the bureaucrats have greater powers and are
appointed from a professional national cadre, while the politicians are drawn from local
communities and have fewer formal powers. Other di erences, in integrity, altruism, per-
sonality, and public service motivation, are areas with fewer priors and more to learn from
the data. We nd that politicians report greater altruism when compared to bureaucrats.
However, while altruism among bureaucrats matters for service delivery, it is the integrity
of politicians that is associated with improved services (not their reported altruism or pub-
lic service motivation). And it is political integrity which is in short supply: politicians
score signi cantly lower than bureaucrats, and those who win elections tend to score less
than contenders who lost.
   Our results suggest that variation in personality of politicians plays a smaller role than
variation in personality of frontline service providers as found in the work of others (Do-
nato et al, 2017; Callen et al, 2015). Traits like conscientiousness and neuroticism might
matter much less for politicians than service providers because the latter are engaged in
day-to-day active service delivery. Integrity, on the other hand, has been highlighted in
theory as a particularly important characteristic of politicians (Bernheim and Kartik, 2014).
However, while political integrity is revealed by the data as the most robust correlate of

                                                6
service delivery, to the exclusion of available measures of personality, altruism, or pub-
lic service motivation of politicians, this could be because these other characteristics are
harder to measure and what is available has not been fully validated in developing country
settings (Laajaj and Macours, 2017). Our contribution is the tip of the iceberg in this regard,
and a necessary one to build empirical evidence on political and bureaucratic selection in
developing countries.
       The empirical signi cance of political quality that we document, together with the
existing theory of public sector bureaucracies which it broadly supports, suggests consid-
erable scope for improving institutional design for local service delivery by taking political
selection into account. Our results thus have implications for how to think about public
policy in “deliverology” (Barber et al, 2011), or improving state capacity to implement pub-
lic policies. In low resource environments, such as in poor countries, it may be possible
to design policies for greater e ect by leveraging the available human resource of local
politicians. Most prior work on local state capacity has overlooked the available theory
on the complex organizations of the state, and the role of politicians in it, focusing instead
on technology (eg. Muralidharan et al, 2016), or exclusively on skills development and
training of bureaucrats and service delivery professionals (World Bank, 2017, critiques the
dominance of capacity building in development practice).7
       Institutional space for the emergence of local politicians has been spreading across and
within countries, even where national politics has veered towards authoritarianism (World
Bank, 2016a). More research and policy experimentation are warranted to understand the
quality of these emerging political leaders at the local level, and how they can play a role in
building state capacity. Available research has focused on understanding the incentives of
   7
     A notable exception is the work of Ra er (2017) which evaluates the impact of training local politicians
in Uganda to play a more e ective role in monitoring the implementation of public investment projects.
The results suggest that better trained councilors, who are provided nancial and technical information
about the projects being implemented in their constituencies, can become more active in exerting political
oversight on the projects, but only in districts where the chairperson does not belong to the national ruling
party. Ra er (2017) interprets this pattern as suggestive of weaker political incentives to monitor where the
national ruling party is more powerful. Our nding of a negative correlation between voter support for the
national political party’s presidential candidate and health service delivery is consistent with Ra er’s (2017)
  ndings. Our contribution of examining the role of political selection goes beyond incentives and supports
policy directions to encourage better quality citizens to enter local governments as political leaders.



                                                      7
politicians to perform, to the neglect of understanding selection. For example, Grossman
and Michelitch (2017) is a recent study from Uganda, where research e orts were devoted
to evaluating a transparency intervention targeted at strengthening political incentives.
Interestingly, the study nds no e ect on service delivery, and the authors explain this
precisely by appealing to the role of multiple actors in service delivery organizations. A
few studies are beginning to examine whether transparency interventions can be designed
from the start to impact political selection, nding that entry of candidates is quite respon-
sive to such interventions, resulting in greater representation of non-elite citizens (Gulzar
and Khan, 2017; Banerjee et al, 2017). Much more work remains to understand whether
such transparency interventions result in the selection of particular political traits, whether
these traits matter for the performance of public sector organizations, and through what
mechanisms.
    The rest of the paper is organized as follows. Section 2 describes the institutional con-
text and data gathered in Uganda to examine the role of selection traits of bureaucrats and
politicians in the productivity of the district organization. This section presents evidence
of systematic di erences between bureaucrats and politicians across several dimensions
of selection. Section 3 provides the empirical speci cation to examine the correlation be-
tween district productivity and selection traits of bureaucrats and politicians. Section 4
presents the main results of the robust correlates of local productivity in delivering health
services. Political integrity emerges as a particularly signi cant and substantial covariate
even after accounting for bureaucrat selection, and other socio-economic-political factors
of the district. Section 5 examines the correlates of political integrity—what conditions are
associated with the selection of better quality politicians. Section 6 discusses the pattern of
evidence and provides an interpretation of the mechanisms behind it. Section 7 concludes.



2    Institutional Context and Data

The district government in Uganda exhibits institutional characteristics that are typical of
the local organizations that deliver public services and implement public policies in devel-


                                              8
oping countries. Bureaucrats appointed to the district by national ministries are responsi-
ble for managing national scal transfers to deliver public services across sectors–health,
education, agriculture, roads, water, environment and natural resources. Following a wave
of political decentralization across the developing world in the 1990s, Uganda promulgated
The Local Government Act of 1997 to provide space for locally elected politicians in the
functioning of the district bureaucracy. District councils, consisting of councilors elected
from sub-district constituencies and headed by a directly elected chairperson, have legisla-
tive functions over matters devolved to the local council by the national state. Most impor-
tantly, these locally elected politicians are expected to perform unique roles of oversight or
monitoring the bureaucracy, and of facilitating service delivery, as a liaison between local
citizens and the state (Grossman and Michelitch, 2018; Ra er, 2017). An original survey
was undertaken between September and December 2015 in 75 of Uganda’s 112 districts (at
the time) to measure the selection traits of these district bureaucrats and politicians who
oversee the implementation of national programs.8 Figure 1 provides a map of our study
area.
       In each District, all the following elected politicians were interviewed: the District
Chairperson or Municipality Mayor; three District/Municipality Woman Councilors and
three District/Municipality Directly Elected Councilors. In addition, all non-elected con-
tenders for the position of District Chairperson were listed and approached for interviews.
In total 770 district politicians and political contenders were interviewed. Among bureau-
crats, all District Chief Administrative O cers (CAO), deputy CAO and the Resident Dis-
trict Commissioners (RDC) were interviewed. An extensive roster of senior o cers were
also included on the sample to be interviewed 9 ; whenever they refused to be interviewed
or the position was vacant, they were replaced by other members of the same department.
A total number of 1,357 bureaucrats are included in the nal sample.
   8
     Annex 9.8 provides the list of surveyed districts
   9
     They are the Chief Finance O cer, Senior Finance O cer (Expenditure), Senior Accountant (Accounts),
District Planner, District Education O cer, Senior Education O cer, Senior Inspector of Schools, District
Internal Auditor, District Production O cer, Senior Agricultural O cer, District Health O cer, Assis-
tant District Health O cer (Environmental Health), and Assistant District Health O cer (Maternal Child
Health/Nursing).



                                                   9
       Our survey respondents thus comprise key o cials in leadership positions in local gov-
ernment, with two distinct lines of selection: (1) local politicians who contest elections and
are voted in/out by locally-residing citizens; (2) bureaucrats who are appointed by national
political leaders and technical ministries. We asked a series of questions in our survey about
who has leadership powers in local government, and found evidence of an informal shar-
ing of power between locally elected politicians and bureaucrats appointed by higher-tier
politicians. For example, we asked: “Who in your district/municipality has the greatest
ability to get civil servants to follow their orders?” Among the appointed bureaucrats and
technocrats responding to this question, 93 percent indicated that the Chief Administrator
O cer (CAO), who is appointed by national politicians, has the greatest authority. How-
ever, among the politicians responding to this question, a lower proportion, 70 percent,
indicated the CAO, and 24 percent indicated that the directly elected District Chairperson
or Municipality Mayor has the most authority. Local journalists—who head the radio sta-
tions which comprise the main news media serving these districts—were even more split
in their responses: 43 percent indicated the CAO, 24 percent indicated the directly elected
local politicians, and 21 percent indicated the Resident District Commissioners, who are
another cadre of senior bureaucrats appointed by national politicians. This setting thus cor-
responds to the special features of public sector organizations that was discussed in Dixit
(2002)–multiple stake-holders (principals) who together manage a complex set of policy
implementation and service delivery activities. The key distinction between the two types
of local leaders is that one type, the politician, is selected locally from among the residents
of the local government through the electoral process, while the other, the bureaucrat,
consist of career technocrats selected by national political and bureaucratic leaders from a
national pool of candidates.10
       Survey modules were developed on the basis of the available literature to measure the
integrity, altruism, cognition, risk-aversion, personality traits, and public service motiva-
  10
      As we will see further below, this distinction can help us understand the patterns revealed by the data
about why it is political selection that appears more signi cant than bureaucrat traits. Bureaucrats may be
strategically appointed by national leaders, making bureaucrat traits more endogenous to district character-
istics which directly predict service delivery.



                                                     10
tion of these bureaucrats and politicians, and are discussed in the section below. To our
knowledge, this is the most comprehensive data gathered on a large number of key actors
interacting in the complex organization of the public sector.


2.1       Selection traits of bureaucrats and politicians

Detailed modules measuring individual characteristics and behavioral traits were admin-
istered for both politicians and bureaucrats. Basic demographic characteristics include
gender, age and educational attainment, as well as household-level characteristics such as
family size and asset ownership. Other than basic demographic characteristics, our survey
includes rich information about individuals’ integrity, cognitive ability, risk preferences,
non-cognitive traits and motivation to work in the public sector. Detailed information
about each of these variables and the construction of aggregate measures are provided in
Annex 9.1 ; here we provide a brief description of these behavioral measures.
       The integrity of respondents was assessed through the Moral Disengagement Measure,
developed by Moore et al. (2012). The Integrity Index is a z-score using responses from
eight sub-components. Taken together, the eight questions presented in Annex 9.4 have
been shown to be correlated with unethical behavior in lab experiments. Individuals are
asked to rate their agreement with the statement on a scale of 1 (Strongly Disagree) to 5
(Strongly Agree). We recode the variables such that higher values correspond to disagree-
ment with unethical behavior, then proceed to take a simple average of the eight questions
and normalize it to obtain the Integrity Index.
       In assessing non-cognitive ability, the survey follows the World Bank STEP (Skills To-
wards Employability and Productivity) methodology. It includes not only the popular Big
Five Index - a set of 15 questions used to assess ve broad domains of personality -, but
also the more speci c dimensions of Grit, Decision-Making and Hostility Bias. The moti-
vation to work in the public sector is measured using Perry’s PSM Index, composed of 36
questions measuring motivation in six di erent dimensions - Attraction to Policy Making,
Commitment to Public Interest, Social Justice, Civic Duty, Compassion and Self-Sacri ce11 .
  11
       Perry’s PSM Index has been widely used in the recent personnel economics literature, such as Dal Bo,

                                                     11
    Risk Preferences are assessed using a simple, non-incentivized risk game, detailed in the
Annex 9.7. Similarly, Altruism is measured by a hypothetical game in which the individual
is given UGX 50,00012 and has to decide how much of it to share with an anonymous
individual.
    To assess cognitive ability, the survey uses a digit span memory test, in which an in-
creasingly longer string of numbers is read to the respondent, who is in turn asked to repeat
it. We use the number of digits the respondent can correctly recall as a proxy for cognitive
ability.13 . We complement this measure, following Dal Bo, Finan and Rossi (2013), with an
indicator assessing whether the respondent chooses a dominated option in the risk game
described above.
    Tables 1 and 2 present descriptive statistics on the survey sample, highlighting di er-
ences in means between politicians (including non-elected contenders) and bureaucrats.
Table 1 presents basic socio-demographics characteristics and con rms the intuition that
politicians are very di erent from bureaucrats in basic observable characteristics. Politi-
cians are much more likely to be females, to have been born in the same District where
                         14
they currently work           and also to come from less wealthy households (they have larger
families and hold less assets). The most striking distinction between bureaucrats and politi-
cians is in educational attainment: whereas half of politicians have not completed college,
96 percent of bureaucrats did, including half with postgraduate degrees. These numbers
are broadly consistent with Ra er’s (2017) data on subcounty politicians and o cials in
Uganda, and paint broadly the same scenario of policy environment: local politicians over-
see project execution led by highly educated and centrally appointed bureaucrats.
    Table 2 presents descriptive summaries of selection traits. Panel A reports, for each of
the components on the Integrity module, the share of individuals who agree or strongly
agree with each statement. For all components a larger share of politicians than bureau-
crats agree with statements such as "People shouldn’t be held accountable for doing ques-
Finan and Rossi (2013) and Callen et al. (2015)
   12
      Approximately USD 15, at the conversion rate of USD 1 = UGX 3,400 (average during the survey period).
   13
      This is the same methodology used in Weaver (2016).
   14
      Which is not surprising given that bureaucrats are centrally appointed.



                                                    12
tionable things when they are just doing what an authority gure told them to do". It’s
worth noting, however, that for all items a signi cant majority of all respondents do not
agree with the statements. Overall, politicians score on average 0.32 standard deviation
lower than bureaucrats on the aggregate Integrity Index.
       Panels B and C present results on personality traits. Politicians do fare better than bu-
reaucrats in other non-cognitive dimensions. They score 0.12 s.d. higher on average in the
Big-Five Index, though their advantage is restricted to two domains in which we would
indeed expect politicians to perform better: Extraversion and Agreeableness. When we
consider other measures of non-cognitive ability, bureaucrats fare better on Grit, Decision
Making and Hostility Bias, such that we nd no distinction between the two groups when
pooling all dimensions above in a Non-Cognitive Index. We nd no signi cant di erence
between politicians and bureaucrats in the Public Sector Motivation Index: bureaucrats
report higher Compassion but score lower on Civic Duty and Self-Sacri ce; in other di-
mensions we observe no gap.15 .
       Finally, Panel D presents results on cognitive traits and measured attitudes of altruism
and risk preferences. Consistent with large di erences in educational attainment, politi-
cians fare signi cantly worse in both cognitive measures: on average they can recall 0.3
fewer numbers than bureaucrats, and are also 3 p.p. more likely to choose a dominated op-
tion on a simple risk game (choosing a UGX 1 million reward instead of a gamble of equal
probability between UGX 1 million and UGX 2 million)16 . Regarding altruism, politicians
donate a slightly larger fraction of a hypothetical prize on average, and also make less risky
choices in lottery games.17
  15
      One advantage of using the same modules as other surveys in the literature is comparing the results
obtained in Uganda with other settings. This paper uses the same modules for measurement of PSM and
Big-Five as Dal Bo, Finan and Rossi (2013) in Mexico, so we can actually compare some of the descriptive
statistics to their Table II. Our mean values for Big Five measures are across the board lower, indicating a
"worse" pool of individuals in terms of these non-cognitive traits. For PSM, however, the scores are pretty
similar.
   16
      The share of individuals who make a mistake, for both politicians and bureaucrats, around 10 percent
in the Uganda sample, stand in stark contrast to the results of Dal Bo, Finan and Rossi (2013) where almost
40 percent of individuals who were being considered for the position of social workers in Mexico made this
error.
   17
      While all gures discussed in these tables refer to di erences in means, Figures 3 and 4 presents the CDF
of traits for each group and Kolmogorov-Smirnov tests reject equality of distributions.



                                                     13
       While Table 2 suggests politicians have very distinct behavioral traits from bureaucrats,
it is possible that all these di erences disappear when we compare individuals with similar
observable characteristics. Is that the case, or do individuals who select to be politicians
fundamentally di er from those at bureaucratic positions once we control for observable
traits such as gender, education and age? Figure 2 presents the average di erence in traits
between politicians and bureaucrats, once we account for a range of observable individual
characteristics. Most of the previous results stand: politicians still perform worse in In-
tegrity and Other Non-Cognitive measures (Grit, Decision Making and Hostility Bias), but
better in the PSM, Big Five and Altruism. The main distinction from the previous ndings
is that di erences in cognitive attainment become very small and statistically insigni cant,
suggesting that any di erences between the two groups are simply mediated by di eren-
tial educational attainment.18


       Consistency of individuals’ traits measures
       One possible caveat to this discussion is the reliability of these behavioral measures,
which are self-reported. Whereas we cannot rule out entirely that desirability bias domi-
nates our measurement, we provide some evidence they seem to capture important dimen-
sions of personality.
       First, as discussed above, there’s quite a lot of variation in answers. Even for questions
that might induce a strong desire to conform, such as "Is it OK to take credit for others’
ideas?", 20 percent of respondents answer they disagree and only 20 percent assert they
strongly agree. Furthermore, we identify large and systematic di erences between answers
of politicians and bureaucrats.
       Second, Table 3 provides pairwise-correlation between several of the individual traits’
measures. All indices are constructed in such way that higher values are considered posi-
tive traits - accordingly, all pairwise correlations (with exception of PSM-Cognitive, which
has a negative point estimate but very close to zero) are positive and most of them sig-
  18
     Indeed simply controlling by education makes the di erence in cognitive attainment become not statis-
tically signi cant (not reported).



                                                   14
ni cant. Personality traits - Big Five, Non-Cognitive, PSM and Integrity - are all strongly
and signi cantly correlated with each other. The correlation of PSM/Big Five is 0.35 and
PSM/Non-Cognitive is 0.39.19 This suggests that these measures, captured in di erent ques-
tionnaire modules, do capture some underlying behavioral traits of the interviewees.20


2.2       Measure of productivity of district bureaucracies

These survey data on bureaucrats and politicians are matched to government data on the
productivity of districts in implementing national programs. The Ministry of Health (MoH)
in Uganda compiles detailed information on the performance of districts in delivering na-
tionally mandated public health programs. Using the Health Management and Information
System (HMIS), the MoH compiles the data reported by local health facilities into Health
District League Tables (DLT) that are intended to capture the relative performance of ser-
vice delivery across districts (See Figure 12 for details). The DLTs are part of the Annual
Health Sector Performance Report by the MoH. The closest year to our survey for which
the DLT data are available is FY 2014/2015. The bureaucrats and politicians we interview
in September-December 2015 were the presiding o cials during FY 2014/2015 when the
health performance indicators were compiled by MoH .
       Table 4, top panel, presents descriptive statistics on the quality of health services across
districts. Whereas there is little variation in some of the individual indicators (the median
coverage of DPT3 vaccination is 98 percent), other indicators vary widely between Dis-
tricts: the average share of women having 4 or more ante natal care visits is only 34 per-
cent (s.d.: 10.9 ) and the average share of pregnant women receiving preventive malaria
treatment in pregnancy (IPT2) is 53 percent (s.d.: 11.6). The DLT total score is a weighted
average of 14 quality care and management practice indicators ranging from 0 to 100 - the
average District scores 73 points. In the analysis below we will often use a normalized ver-
sion of the total score, so that changes can be interpreted as standard deviation variation
  19
      Dal Bo, Finan and Rossi (2013) report a 0.54 correlation for PSM/Big Five. Whereas we nd a rather pre-
cisely estimated zero correlation between PSM and our Cognitive measure, Dal Bo et al nd a 0.12 correlation
between PSM and IQ.
   20
      In Annex 9.2 we provide interitem average correlations and Cronbach’s alphas for both our aggregate
indices and its components.

                                                    15
in the score (7.25 points).
    These health service delivery indicators are the best measures we have available in
our study context to capture the productivity of the district bureaucracy as a complex
organization. Rasul and Rogger (2016) and Rasul, Rogger and Williams (2017) review
the available measures of productivity of national bureaucracies and focus on completion
rates of projects undertaken. The advantage of the setting we examine–of district-level
bureaucracies–is that productivity can be measured in terms of services that are directly
linked to development outcomes, such as health in poor communities.



3    Empirical Speci cation

The empirical analysis in this paper is not positioned to causally identify the e ect of any
variable on health service delivery performance. Rather, we examine what are the robust
correlates, if any, of variation in performance across districts using all available data on
a variety of potentially signi cant correlates for this purpose. The main question guiding
the analysis is whether (which) selection traits of bureaucrats and politicians are signif-
icant in explaining district performance. In order to control for the role of other district
socio-economic and political characteristics in driving any correlation, by simultaneously
in uencing both selection traits and health service delivery, we use other District level
administrative data on these district characteristics.
    Speci cally, we estimate variants of the following equation:


                                        ¯ β + Z θ + γj +
                              Yij = α + X                      ij                           (1)
                                         i     i




    where Yij will often be the standardized health score in the District League Table for
district i in region j. In some instances, Yij will instead be speci c components of the DLT
                  ¯ i includes our main variables of interest: the average trait of politicians
score. The vector X
and/or bureaucrats in District i; di erent speci cations will use di erent subsets of traits.
Zi is a vector including a series of socio-economic and political characteristics of districts


                                              16
and γj allows for region-speci c intercepts.
       The variables contained in the vector Zi come from administrative and survey data. The
2014 Census provides information on Districts’ population and urbanization rates, while
poverty rates are estimates for 2013 (World Bank, 2016b)21 . We complement these socio-
economic data with average students’ score at the UCE in 2011, a nationally administered
exam students take at the end of lower secondary school, as the best available proxy for
education in the population.22 Our survey also gathered data on the population of radio
stations operating in Uganda at the time of the survey, from which we derive a variable
of the number of radio stations headquartered in a district. Access to radio, and to media
in a more general sense, has been shown to impact the public scrutiny of local politicians
(Ferraz and Finan, 2008) and a ect citizens’ investment in health and education (Khemani
and Keefer, 2015; Khemani and Keefer, 2014). We also consider that the existence of local
radio stations might proxy for other unobservables such as the dynamism of local markets.
       In addition, we compile data on local vote shares from the 2011 elections, the last one
before our survey. The main variables are the vote share of the presidential candidate of
the NRM, the national ruling party (labeled "Voter Attachment to National Party"), at the
district level, and voting Her ndhal Indices23 for district politicians 24 .
       The lower panel in Table 4 presents descriptive statistics of administrative data for the
75 Districts included in the survey. The median District in our sample has approximately
250,000 people and is overwhelmingly rural25 . Poverty rates, at 23 percent in the median
District, are only slightly higher than the national rate of 20 percent in 2013. The numbers
also show the subdued level of political competition in presidential elections: the candidate
  21
      These are small areas estimates produced by the World Bank, using data from the 2013 UNHS and 2014
Census data. For details on methodology see for example Ebers, Lanjow and Lanjow (2002).
   22
      Note that higher values of the UCE score are indicate of lower education performance, as described in
the data Annex. We include 2011 UCE performance since those cannot be in uenced by current politicians,
who were elected in that year.
   23
      The Her ndahl Index is most commonly used as measure of rm competition in a given market, but has
been widely used to measure political competition . The Her ndahl Index is obtained, at the electoral area
level, by the sum of the squares of candidates’ votes. Higher values are associated with a less competitive
environment (the index ranges from 1/N to 1, where N is the number of candidates).
   24
      We compute all the results at the electoral area level and then obtain the district-level average across all
areas.
   25
      According to the 2014 Census only 16 percent of the population in Uganda lived in urban areas.



                                                       17
of the ruling party won 70 percent of the popular vote for president in the median district
during the 2011 presidential election.



4    Results: correlates of district productivity

This section presents the main results of the paper, assessing the correlation of politicians’
and bureaucrats’ selection traits and service delivery. We start by presenting raw corre-
lations between actors’ selection traits and districts’ health outcomes. Here we document
that integrity is the only trait of politicians’ that is correlated with health outcomes, while
among bureaucrats’ both risk-aversion and altruism show positive correlations with health
quality.
    Our main result is to assess whether these correlations hold in multivariate regressions.
Because of the problem of having a small number of observations (75 districts) relative to
a large number of potential regressors (40 in total, counting all selection traits and district
socio-economic and electoral characteristics on which we have data available), we face
the issue of model selection – should we just include all covariates, at the cost of having
very noisy parameter estimates, or should we try to restrict the model to covariates we a
priori think are “better”? The available economic theory provides little guidance on this.
Our reading of the literature suggests that a machine-learning process such as LASSO has
become the methodology of choice in such situations, to allow the data to reveal which
variables have greater predictive power. We therefore use a model that includes variables
chosen by LASSO – which drops variables that explain little of the variation in the data–
as our main speci cation. We think this strategy should make our readers less worried
that we are “cherry-picking” a model that makes some results look better. We also present
in the robustness section a range of OLS estimations with an exhaustive list of controls.
These estimates suggest that politicians’ integrity is the most robust predictor of variation
in health outcomes among all selection traits of policy actors.
    We then proceed to perform a series of additional robustness checks on our main result.
Having established the role of politicians’ integrity as an explanation for overall health


                                              18
service quality, we show that the correlation holds not only for the aggregate score but also
for a majority of individual sub-components of the index. The results become stronger if
we use the rst principal component of individual health indicators.
       Finally, we provide evidence consistent with the mechanism that higher political in-
tegrity in uences bureaucratic productivity through improved oversight and facilitation
by district councilors who are elected at the frontlines of health service delivery. Among
the three types of politicians surveyed - local councillors, District Chairperson and non-
elected Contenders -, it is the integrity of councillors, who are supposed to monitor and
facilitate delivery, that is robustly correlated with several indicators of health service de-
livery.


4.1       Overall correlations

To investigate whether characteristics of politicians and bureaucrats in Uganda’s districts
are robustly correlated with the quality of service delivery, we start by presenting simple
graphs showing how the DLT Health Scores and average traits of local policy actors covary
in the sample of 75 districts in Figures 6 and 7. To account for broad di erences in income
across districts, we regress the DLT scores on district poverty rates, and present the scatter
plot of estimated residuals against selection traits.
       The rst nding is that, among politicians, integrity is the only robust predictor of
variations in DLT between districts: a 1 s.d. increase in politicians’ average Integrity Index
is correlated with a 0.3 s.d. increase in the residual of health score. The only other trait
of politicians that is marginally correlated with health scores is Altruism. Importantly,
personality traits like those measured by the Big Five Index and Public Sector Motivation
are not correlated with quality of service delivery.
       Among bureaucrats, the scatter plots indicate no correlation between average integrity
and quality of service delivery. Three traits of bureaucrats are signi cantly correlated with
DLT scores: districts where bureaucrats are more risk-averse and more altruistic, and Dis-
tricts where bureaucrats’ PSM is lower, also have higher scores.26 .
  26
       The correlation between Risk-Aversion and health scores, however, disappears if we drop the outlier

                                                     19
    While these scatter plots are informative of broad correlations, individual bureaucrats’
and politicians’ traits most likely covary with other District level characteristics such as
education levels and support for the central government. In order to assess how robust are
these correlations when we compare similar Districts, we estimate di erent speci cations
of Equation (1), rst including each one of the average traits of bureaucrats and politicians
                               ¯ i is always a scalar). Figure 8 reports point estimates and
separately as regressors (i.e. X
CI for the coe cient on average traits. The results are broadly consistent with the simple
correlations. Among politicians’ traits, only the average Integrity Index is signi cantly
correlated with better health outcomes, suggesting that a 1 s.d. increase in average integrity
is correlated to 0.23 s.d. increase in the DLT score 27 . Regarding bureaucrats, altruism is
the only trait that is (marginally) signi cantly correlated with health outcomes once we
control for District level characteristics. These results suggest that, among all behavioral
traits measured for politicians and bureaucrats, politicians’ integrity seems to be the one
robustly correlated with service delivery - even after taking into account several other
socio-economic characteristics28 .


4.2     Assessing predictive power of all behavioral traits

While the previous section reports ndings from regressions that include each selection
trait separately in equation (1), we would like to assess whether the ndings hold when
we allow for all individual traits to a ect health outcomes simultaneously. Given the large
number of individual traits we measure in the survey and the limited number of obser-
vations (districts), we are faced with the question of what model to estimate: should we
include all individual traits, at the expense of lots of imprecision in estimates? Should we
restrict attention to politicians’ integrity and include districts’ characteristics as controls?
    Following the literature, we proceed by reducing the "degrees of freedom" in model
selection and instead relying on regularization techniques of machine learning 29 (Athey,
district of Amudat.
   27
      The estimated e ect is 30% smaller than in the simple correlation, suggesting that part of the observed
raw correlation is driven by other district characteristics that covary with politician integrity.
   28
      Results in regression form are presented in Table A6
   29
      We nonetheless report OLS estimates of di erent speci cations as robustness checks in Table 5.

                                                     20
2018). In particular, we use the LASSO procedure (Tbishirani, 1995) to choose, among all
District characteristics and interviewees’ selection traits, which variables are most predic-
tive of DLT scores. LASSO is considered to be particularly suited, among other machine
learning techniques, when estimates for several of the potential explanatory variables are
believed be approximately zero (Abadie and Kasy, 2017). In practice, the LASSO minimizes
an objective function that includes a goodness-of- t term (sum of squared errors) and a
penalization for the sum of covariates’ coe cients. The weight of the penalization, the
lambda parameter, is chosen by cross-validation, i.e. by estimating the model in test sub-
samples and evaluating its performance in the remainder of the sample (Athey, 2018).
       Following Best, Hjort and Szakonyi (2017), we proceed in two steps. We rst estimate
a LASSO model30 with District-level health score as the dependent variable and a set of 40
covariates. We then select the variables with non-zero coe cient in the LASSO model and
report the results of both bivariate and multivariate regressions.
       The 40 variables used in the LASSO model include the average for 12 selection traits
already explored, separately for bureaucrats and politicians: PSM, Integrity, Big Five, Non-
Cognitive, Altruism, Indicator for mistake in risk game, Cognitive, Risk Aversion, Assets
and indicators for 3 education levels. We also include regional dummies and the following
socio-economic and political District level characteristics: vote share of the presidential
candidate of the NRM, Number of radio stations, Poverty rates, log Population, Share of
urban population, Her ndahl index for Councillors vote share and the UCE score as a
measure of education in the district population (lower values are asdociated with better
education outcomes).
       The LASSO model selects eight variables with non-zero coe cients: four socio-economic-
political characteristics of Districts (Vote share of the presidential candidate of the NRM,
Poverty Rate, Number of Radios and UCE score), three bureaucrats’ traits (Altruism, Risk-
aversion and indicator of less than college education) and politicians’ Integrity. Figure 9a
reports the point estimates and 95 percent con dence intervals for bivariate regressions
  30
     Speci cally, we use Friedman, Hastie, and Tibshirani (2010) algorithm implemented in the R-package
-glmnet-. It runs a LASSO regression with a lambda parameter that minimizes the 10-fold mean cross-
validation error.


                                                  21
for each of the selected variables, while Figure 9b reports the results for a multivariate
regression including all selected variables 31 . All variables are standardized, so the coe -
cients can be compared and interpreted as the e ect of a 1 s.d. increase of the explanatory
variable on the health score.
       The integrity of politicians appears to be the most relevant selection trait to a ect health
scores: in the multivariate regression, we estimate that an increase of 1 s.d. in average in-
tegrity is correlated with a 0.24 s.d. increase in the Health Score, while all other point
estimates for selection traits are smaller in absolute size (though not statistically di erent
from the Integrity estimate) and sometimes not statistically di erent from zero. The esti-
mates also suggest that bureaucrats’ altruism signi cantly impact health outcomes, but the
coe cient is half the size of politicians’ integrity. The indicator for bureaucrats with lower
than college education (recall that only 4 percent of bureaucrats have less than college
education) is also large in magnitude (-0.2 s.d.) and signi cant.
       It is worth noting that the only variables with larger absolute e ect are the number of
radio stations headquartered in a district and voter attachment to NRM: an increase in 1
s.d. in the number of radio stations is correlated with a 0.3 s.d. in the health score; while
a 1 s.d. increase in vote share for the NRM presidential candidate is correlated with a 0.25
lower health score. 32 .


4.3       Robustness

We start our robustness checks by estimating OLS regressions with Districts’ DLT score
as dependent variable and politicians’ average integrity as explanatory variable in Table 5,
and estimating models with increasing number of controls. Column (1) presents the sim-
ple bivariate regression, while column (2) reproduces the estimate in Figure 8, suggesting
a 1 s.d. increase in average politicians’ integrity is correlated with a 0.23 s.d. increase in
the health score. In column (3) we include not only districts’ characteristics but also other
selection traits of politicians as explanatory variables. The coe cient on politicians’ in-
  31
    Full results in table format are presented in Table A8
  32
    Note that if individual traits are measured with more noise than Districts’ overall characteristics, we
should expect the coe cient on traits to be biased toward zero.

                                                    22
tegrity barely changes, but standard errors become larger and the coe cient is no longer
statistically signi cant (p-value = 0.11). As in the model chosen by LASSO, we nd that
a higher number or radio headquartered in a district is a strong predictor of better health
services, as is scores at the UCE national examination.
   In column (4) we add the average integrity of bureaucrats as explanatory variable in
order to assess if it has independent explanatory power. The coe cient on politicians’
integrity remain unchanged, and that of bureaucrats’ integrity is an order of magnitude
smaller and not statistically di erent from zero. Given the large standard-errors of esti-
mates, however, we cannot reject that these two coe cients are equal (p-value = 0.41). In
a very demanding test, in column (5) we include all of bureaucrats’ and politicians’ traits
simultaneously as explanatory variables, together with districts’ characteristics. In this
model, the coe cient on politicians’ integrity is smaller (0.16) and not statistically di er-
ent from zero. We should caution, however, that in this speci cation we are estimating 40
parameters with 75 observations. That is, the F test of a “horse race” between bureaucrat
and politician integrity, is a weak test in our setting. We provide other tests below that re-
veal politicians’ integrity, not that of bureaucrats’, as a robust correlate of service delivery.
The available economic theory does not provide any predictions about whether bureaucrat
or politician characteristics should matter more. We do not conclude from the results that
bureaucrat traits don’t matter, and indeed nd that bureaucrats’ altruism does show-up as a
signi cant correlate in several speci cations. Other bureaucrat traits might matter as well,
but perhaps we are unable to discern it for a variety of reasons (we are underpowered; we
have not measured the relevant traits well; local bureaucrats may be strategically appointed
by national bureaucrats and politicians, making bureaucrat traits more endogenous to dis-
trict characteristics which directly predict service delivery.) The one result that is revealed
systematically by the data is that political quality (and speci cally, the characteristic of
integrity) matters signi cantly for the productivity of bureaucratic organizations.
   All results presented so far document how politicians’ integrity correlates with overall
health scores. The details of the health service delivery index construction are presented
in Annex 9.3, but it’s reasonable to assert that the weights attributed to each of the com-

                                               23
ponents, while linked to overall policy objectives, are somewhat arbitrary. This section
decomposes the overall index in its individual indicators and evaluates the correlation be-
tween politicians’ Integrity and each of the components. Figure 13 reports regression esti-
mates of the coe cient on politicians’ integrity, using the same speci cation as in column
(2) of Table 5, but having each of the sub-components indicated as dependent variable. We
standardize all components so that the coe cients are comparable across regressions.
       The results reported in Figure 13 suggest that Politicians’ integrity is robustly positively
correlated with the majority of the components: the coe cient is statistically di erent
from zero (at 10 percent level) for eight of the fourteen components, and point estimates
are positive for all but four components. It suggests that an increase of 1 s.d. in integrity
is correlated with a 0.4 s.d. increase in anti-Malaria coverage for pregnant women (IPT2
coverage); a 0.4 s.d. increase in the share of households having a latrine; and a 0.3 s.d.
increase in the share of infants delivered in government or Private Non- Pro t facilities.
Higher politician’ integrity is also positively correlated with measures of good administra-
tive practice in the health sector, such as completeness and promptness of health facilities’
reporting.
       In a further robustness test on how much our nding depends on the measure of health
outcomes, we replace the DLT score as dependent variable with the rst component of a
Principal Component Analysis (PCA) of the 14 health indicators. The rst principal compo-
nent used explains 22% of the variance in the sample 33 . We replicate the OLS speci cations
from Table 5 in Table 6, but using the rst component as dependent variable. Results are
broadly similar: estimates suggest that an increase of 1 s.d. in average politicians’ integrity
leads to an increase of 0.2-0.4 s.d. in the rst component of health outcomes. Note that here,
using the rst component of health outcomes instead of the health score, the estimated co-
e cient on politicians’ integrity remain statistically di erent from zero in all speci cations,
even after including all of politicians’ and bureaucrats’ traits simultaneously. Furthermore,
  33
    It’s worth noting that the PCA has negative factor loading in four our of our fourteen variables: TB
success rate, Completeness of monthly reports, Completeness of facilities reports and Reports sent on time.
This means that higher values in these variables, which should re ect higher performance, has a negative
e ect on the rst component.



                                                    24
we reject in the speci cation in column(3) that the e ect of bureaucrats’ integrity is the
same as politicians’ integrity (p-value < 0.1).
       We also investigate whether the correlation between politicians’ integrity and health
outcomes holds when we measure integrity for di erent groups of politicians. Table 7
presents results where the explanatory variable is the average Integrity Index of di erent
groups of politicians. Column (1) replicates the result that, once we control for several Dis-
trict characteristics, a 1 s.d. in integrity is associated with a 0.23 s.d. increase in the health
score. Column (2) through (4) assess how this result changes if we calculate the average
integrity for each politician type: local councillors, chairperson and political contenders.
Results are noisier and in no speci cation statistically di erent from zero. When we in-
clude all three integrity measures simultaneously, nonetheless, it is the average integrity
of councillors that remains predictive: the coe cient is 0.27, very similar to estimates for
the overall politician integrity, and the coe cients on both chairperson and candidates are
much smaller and indistinguishable from zero34 .
       In Figure 11 we repeat a similar exercise as performed for overall integrity, assessing the
correlation between politicians’ traits and each of the subcomponents in the overall health
index. The results suggest that it is indeed councillors’ integrity which drive the correla-
tion with good service delivery: not only councillors’ integrity is signi cantly correlated
with health indicators such as IPT2 and antenatal coverage, but point estimates are often
larger than those obtained using chairperson’s or contenders’ integrity measures. These
results, disaggregated by type of politician, are somewhat sensitive to including health
spending and age of district as controls– the estimates of the correlation with councillor’s
traits becomes noisier for some of the individual health indicators, while the estimates of
correlation with chaiperson’s traits become more signi cant.35
       As a nal check of robustness, in Table 8 we examine whether other organizational
  34
      Note, nonetheless, that in this speci cation we lose 10 observations since we do not have chaiper-
son/candidate observations for all districts.
   35
      These results are not reported in the paper because including health spending reduces sample size by 3
observations, owing to missing values for 3 districts. The overall signi cance of average political integrity
is unchanged. We discuss further in the next section that the sensitivity when including age of district is
owing to signi cantly lower integrity of chairpersons in young districts, as erstwhile councillor positions
are converted to chairperson positions with the carving-out of a new district from a councilor’s seat.


                                                     25
characteristics, that have been highlighted in the literature, matter in our data and whether
our main results are robust to including them. First, in column (1) we include as control
an index of bureaucrats’ perception on management quality36 . Rasul and Rogger (2016),
drawing upon Bloom and Van Reenen (2007), have documented a positive correlation be-
tween management practices and project completion in Nigeria. Including a variable on
management practices does not change the positive e ect of politicians’ integrity, and the
coe cient on the management variable is not statistically di erent from zero, although the
point estimate is positive and large.
       We also examine whether ethnic heterogeneity among the districts’ bureaucrats and
politicians play a role in service delivery in columns (2)-(3), as in the literature on the
role of ethnic diversity in teams (Hjort, 2014) and in economic development more gener-
ally (Alesina and La Ferrara, 2005; Alesina et al, 2017). Our proxy for ethnic diversity is a
Her ndahl Index of self-reported ethnicities. We rst include as control the index for all
surveyed individuals, and in column (3) we include separate her ndahl indices for bureau-
crats and politicians. Once again the correlation between integrity and service delivery
remains broadly unchanged. The point estimates on the her ndahl indices are small and
not signi cant. In column (4) we include both ethnic diversity and management practices
measure as controls; the coe cient on politician integrity remains signi cant, and with
little change in size across all these di erent speci cations.
       Finally, we consider whether the integrity of di erent types of bureaucrats–those with
greater leadership powers, such as the Chief Administrative O cer (CAO) and the Resident
District Commissioner (RDC), versus those within the technical health bureacracy, such as
the District Health o cer– matters for service delivery, and whether this disaggregation
by bureaucrat type changes the results for political integrity. We calculate the integrity in-
dex separately for the leaders of the bureaucratic organization - CAOs and RDCs - and for
bureaucrats working in the health sector, and include these in the speci cation reported
in column (5). The coe cient on bureaucratic leaders’ integrity is very small in magni-
  36
    It includes six questions on whether bureaucrats believe that senior management recruit and promote
the best employees.



                                                  26
tude, while that for health bureaucrats is somewhat larger but noisily estimated. More
importantly, we see no di erence in the results on politicians’ integrity.



5    Correlates of integrity

The previous section documented a robust correlation between politicians’ integrity, in
particular that of local councillors’, and health outcomes. While we are not in a position
to identify an exogenous source of variation in political integrity, we would like to explore
where variation in integrity might be coming from by exploring its covariates in the data.
Our question in this section is therefore: which variables predict high integrity among
politicians?
    We start this analysis by presenting disaggregated data on behavioral traits by type of
politician; do Councillors and elected Chairs, or contenders vs. elected Chairs, di er sys-
tematically from one another? Table 9 presents average characteristics of Chairpersons,
non-elected contenders for Chairperson and Councillors. The rst overall nding is that
Councillors underperform elected Chairpersons and contenders for that position: they are
much less educated, perform worse in the cognitive test and score lower in Integrity and
Non-Cognitive measures. Among those who ran for Chairperson, di erences are less sys-
tematic between winners and losers: winners seem to perform slightly worse in Integrity
and Non-Cognitive measures, but estimates are noisy. The one signi cant and meaningful
di erence is in the asset index, suggesting winners come from wealthier households than
those defeated.
    We then proceed to evaluate whether politicians’ integrity is consistently correlated
with Districts’ socio-economic and political characteristics and individuals’ observable
traits. Column (1) of Table 10 presents estimates for all politicians, while columns (2)-
(4) restrict the sample to Chairpersons, Contenders to Chairperson and Councillors, re-
spectively. Two results are worth noting. One, electoral competition is associated with
higher political integrity among winning candidates; and two, younger districts tend to
have worse integrity among chairpersons. Lower competitiveness, as measured by the


                                             27
average Her ndahl index for councillors, is correlated with worse integrity in the pooled
sample, as well as separately for elected chairpersons and councilors (signi cant at the 10
percent level). Chairperson integrity is lower in younger districts–those that are carved
out of sub-district units which used to be the seats of councillors. This result is consistent
with the general nding that councillors tend to have worse integrity than chairpersons,
as chairpersons in younger districts are more likely to come from the pool of councillors.
Theoretical models of candidate entry into political markets (Caselli and Morelli, 2004;
Mattozzi and Merlo, 2008; Leon, 2013) may be extended to account for the empirical pat-
tern we observe–that politicians in lower-level positions have worse integrity. Politicians
in lower-level positions of councillors face greater challenges in signaling their type to vot-
ers, because their actions and impact on service delivery outcomes are harder to discern,
since they have fewer formal powers over service delivery compared to higher-level actors.
For example, lower-level politicians can claim that service delivery failures are all due to
lack of funding being provided by senior district and national-level leaders. Therefore, fol-
lowing the logic of the theoretical literature, "bad" types may be more likely to enter as
candidates, and win, in lower-level positions compared to higher levels. However, other
strands of literature would suggest that voters have more information about more local
politicians, perhaps because they are closer to their social networks (Munshi and Rosen-
zweig, 2008; Casey, 2015). Our nding of lower integrity among more local politicians is
therefore a puzzle that cannot be easily explained by arguments available in the literature.
   In other results, politicians from more urban Districts score worse on the integrity in-
dex. Older politicians and those with a post-graduate degree score higher, on the other
hand. These results, however, are not very robust when we restrict the sample to di erent
kind of politicians: standard errors of estimates become very large and coe cients often
change signs. Furthermore, estimates for region xed-e ects are not signi cant, suggest-
ing that average integrity of politicians does not vary substantially after accounting for
other observables. Therefore, the evidence suggests that, as far as measured, politicians’ in-
tegrity is not systematically correlated with Districts’ socio-economic characteristics (that
is, characteristics other than electoral competition).

                                              28
   We also examine the correlation between integrity and other behavioral traits. In docu-
menting bivariate correlations between integrity and a wealth of personal characteristics,
we follow Cantone et al. (2016) and present p-values adjusted for multiple hypotheses
testing. Table 11 presents these correlations using the disaggregated components of be-
havioral indices explored above. Both Public Sector Motivation and Non-Cognitive traits
are positively correlated with reported Integrity. In particular, politicians scoring higher
in Commitment to Public Interest, Social Justice, Compassion and Self-Sacri ce (PSM in-
dicators), and Extraversion, Conscientiousness and Decision Making (Non-Cognitive mea-
sures) also score higher on the Integrity Index. Cognitive ability, Altruism or Risk-Aversion
are not signi cantly correlated with integrity; there is also no correlation with gender or
household asset index. While completing college is not a signi cant predictor, having a
post-graduate degree is signi cantly correlated with higher integrity.
   Having shown that several of the behavioral traits are systematically correlated with
higher integrity among politicians, we proceed to evaluate whether these individual cor-
relations hold when we keep xed other characteristics. Table 12 presents the estimated
coe cients of a regression with politicians’ Integrity Index as dependent variable, and also
include District xed e ects. Several of the patterns presented above hold in a multivari-
ate regression setting: measures of PSM and Non-Cognitive ability (Big Five) are strong
correlates of individual integrity. Among measures of public sector motivation, politicians
with higher Commitment to Public Interest and Compassion also score higher on Integrity,
while for non-cognitive measures Extraversion and Conscientiousness feature as promi-
nent correlates.
   The pattern of correlation between integrity and Conscientiousness, and between in-
tegrity and one particular component measure of the Public Service Motivation index–
Commitment to Public Interest– suggests that the in uence on bureaucratic productivity
of political selection is occurring through greater intrinsic motivation to perform assigned
duties (Conscientiousness), as well as, possibly, greater pro-social preferences (Commit-
ment to Public Interest) which motivate agents to deliver public goods such as public health.
Future work on political and bureaucratic selection in local bureaucracies could focus on

                                             29
better measuring pro-social preferences, and attempt to disentangle the roles of di erent
sources of non-pecuniary motivation, whether arising from integrity or conscientiousness
to do one’s job as tasked, or, preferences for public goods. The currently available measure
of Public Service Motivation is drawn from a literature focused on American bureaucracies
and the political context of the United States (Perry, 1996) and needs greater testing and
validation in poor country contexts. As mentioned before, we regard the results we present
here as just a start, albeit an important rst step, towards closing the empirical gap between
theory and empirics on the role of non-pecuniary motivation in shaping productivity of
bureaucracies.



6    Discussion of mechanisms

The interpretation we o er of the robust correlation between political integrity and health
service delivery is that in districts where higher integrity politicians are endogenously
selected, these politicians are more motivated to perform their multiple roles within the
complex organization of district bureaucracies to improve public health, including sensitiz-
ing citizens to the value of health services, and monitoring and supporting health workers
at the frontlines to deliver. As discussed in section 2, government institutions in Uganda
explicitly provide a role to locally elected politicians to monitor and facilitate service de-
livery, working as liaisons between citizens and the technocrats in charge of delivering
nationally-mandated programs. In particular, primary health and education services in
Uganda have been examined in prior research as areas where local politicians exercise in-
 uence (Akin et al, 2005; Reinikka and Svensson, 2004; Brosio, 2000). This prior research
has, in fact, raised concerns about local political accountability for providing public goods
and managing scal resources. For example, Akin et al (2005) argue that district govern-
ments are more likely (than the center) to spend on private goods rather than public goods.
Brosio (2000) refers to opinion surveys of local councilors in Uganda to suggest that these
councilors are more concerned about paying themselves allowances rather than ensuring
that health facilities are well equipped with drugs and implements. In a celebrated study of


                                             30
“leakage” in public spending, Reinikka and Svensson (2004) suggest that local politicians di-
verted funds from a national transfer of capitation grants intended for schools, and instead
used these funds for their political campaigns. In such an environment, higher integrity
politicians can be reasonably expected to di er in their motivation and therefore in the
degree to which they positively in uence service delivery. However, there are competing,
alternate explanations that we consider below.
   Perhaps the most important alternate explanation is that underlying (unobserved) con-
ditions of household demand for health, or social norms that facilitate health service deliv-
ery, may be simultaneously shaping the selection of politicians and service delivery perfor-
mance, with no direct connection between the two. Other explanations centered around
omitted variables, that local economic and political conditions are driving both political
selection and health services, are somewhat attenuated by our ability to control for local
area poverty estimates, education, and electoral competition (as reported in the results sec-
tion). However, citizen demand and social norms are harder to account for. We argue that
two variables–locally headquartered radio stations, and vote share of the national politi-
cal leader–are likely to go a long way towards mopping-up the direct in uence of citizen
demand and social norms. These are, in fact, the only two other variables, other than politi-
cal integrity, that are robustly signi cant in explaining variation in health service delivery.
However, the signi cance of political integrity survives the inclusion of these variables.
This suggests that political selection plays an additional and independent role in shaping
service delivery, through the behavior of politicians. Nevertheless, the coe cient on polit-
ical integrity that we are able to estimate can be biased because of omitted variables, such
that our results cannot be extended to predicting the size of any service delivery impact of
random assignment of politicians of higher integrity.
   What about reverse causality? This is the other problem to consider when exploring
how to interpret a robust empirical correlation in the data. Would places with better health
services be more likely to select better politicians? There is nothing in the voluminous
literature on Uganda that would support such an interpretation. One reason to doubt it is
that the health services being measured are related to maternal and child mortality, rather

                                              31
than to adult human capital, such as through education, which may shape political behavior
such as deciding whether to vote for corrupt or clean candidates. We use data on districts’
average students’ score at the UCE, a nationally administered exam students take at the
end of lower secondary school, as the best available proxy for education in the population.
There is indeed a positive correlation between better education outcomes and better health
service delivery. However, including or excluding the education variable does not change
the coe cient on political integrity.
   Another way to think about reverse causality is consistent with our preferred inter-
pretation of mechanisms (that politicians of higher quality take better actions to improve
service delivery): if experience with better health services gives citizens greater trust or
con dence in local institutions, and thus makes them more likely to select clean candi-
dates, part of the reasons why they would do so is because of the expectation (that should
be ful lled if this were the equilibrium) that clean candidates will be able to do a better
job of ensuring service delivery. There is a speci c role district councilors are supposed
to play to monitor health facilities. If active citizens choose better quality politicians, by
that very logic, we may reasonably expect them to choose politicians who can do a better
job monitoring the health facilities. Hence, the proximate reason behind the correlation of
political integrity and service delivery would, logically, be more active politicians.
   In ongoing work in a seperate research project in the district of Hoima, we discover
qualitative evidence of the role of local politicians (Ferraz, Habyarimana, and Khemani,
ongoing). We held meetings with district bureaucrats and politicians and asked them to
deliberate upon the most important problems of their district and what they could do to
improve the situation. Health services were identi ed by the local politicians in Hoima as
the most important problem, and the role of local politicians was highlighted to empower
technical o cials to deliver it. Interestingly, our data shows that Hoima’s district coun-
cilors have signi cantly higher integrity than councilors in other districts, although they
are undistinguished on other selection traits.
   In the course of the deliberations in Hoima, the District Health O cer is recorded as
saying: "We as technical people, we need support from di erent leaders", to which the directly

                                             32
elected district chairperson responds as follows: "And as leader, I want to promise total
political support to my technical team, where it is necessary. . . political support, we shall
not interfere with what you are doing in line with health. So we should also empower our
health workers, down on ground that you are in charge, feel yourself that you are in charge
and when you’re commanding for example the health in-charge your sta , let them feel that
you are the boss of that health facility and they are supposed to do what you have directed
them to do. Be empowered that you are the boss and you want something to happen as we have
agreed here. That empowerment is what am talking about ". In another part of the discussion,
the same politician talks about the importance of lower-level politicians–the councilors–
monitoring health facilities: "Our parish chiefs must make sure that they are monitoring our
facilities and write reports, the parish chief can report to our sub county chief, then the sub
county chief writes a report to the DHO, then the DHO writes a report to the CAO. So that now
the CAO can present such ideas for action and adjustments where necessary. So that is what
we are now going to plan to put into place. . . to enforce because it is in the existence...". This
exchange suggests that local politicians can play both negative and positive roles in the
delivery of services–interfering to obstruct versus monitoring to facilitate better services.
   The integrity of local politicians can be expected to matter in whether they obstruct
or support the technical o cers in delivering services within the complex organization
of local government. For example, if local politicians are colluding with health service
providers to extract rents from the health facility, they may obstruct disciplinary action by
district management. In another study, in Pakistan, Callen et al (2014) nd that doctors
with connections to political leaders are more likely to be absent from public health clin-
ics, and the public o cials who manage these doctors are more likely to report political
interference when trying to apply sanctions. In a study in Uganda, in the district of Arua
(which is in our sample, and has signi cantly lower average political integrity compared to
Hoima), Grossman, Platas and Rodden (2018), report qualitative evidence that the Medical
O cer In-charge of a health center had been absent for several months, and that it took a
community-wide meeting facilitated by external actors to get this issue to be reported to
the district leaders and prompt them to take disciplinary action. In such a setting, the in-

                                                33
tegrity of local politicians who are supposed to monitor and make reports to management,
can play a direct role in whether the politicians take actions that improve or discourage
service delivery. Grossman, Platas and Rodden (2018) also provide qualitative evidence
that district bureaucratic leaders rely on local councilors to resolve service delivery prob-
lems raised by citizens, since these councilors are located at the sub-district level where
service delivery happens, and are equipped to mediate between citizens and frontline ser-
vice providers. The results of Groassman et al’s (2018) evaluation of a citizen engagement
intervention in Arua shows that these interventions targeted at increasing citizen activism
did not impact health service delivery. This evidence supports our interpretation that cit-
izen activism alone cannot explain the correlation between political quality and service
delivery; actions by local politicians are likely to be the proximate route through which
political quality matters.



7    Conclusion

This paper contributes to understanding local state capacity to deliver public goods. In a
recent review, Dal Bo and Finan (2016) list as an open question the e ect of decentraliza-
tion on the delivery of public goods, despite more than two decades of policy e ort towards
establishing local governments in poor countries. The research agenda has shifted away
from comparing centralized versus decentralized provision of public goods towards exam-
ining principal-agent relationships within myriad complex organizations of government.
Key ndings from available research are in the nature of con rming basic economic prin-
ciples such as: pay for performance tends to strengthen incentives and improve service
delivery (Banerjee et al, 2008); higher wages can improve the selection of state personnel
(Dal Bo et al, 2013); and monitoring can reduce corruption (Olken, 2007).
    The open question which is particularly relevant for policy-makers in Africa and other
developing regions is how to structure principal-agent relationships in frontline jurisdic-
tions of government, that is, the local governments which implement national policies and
deliver spending programs. Local bureaucracies are tasked with multiple complex jobs,


                                             34
such as managing cadres of health workers and teachers to deliver basic health and educa-
tion services on which poor households rely, and implementing infrastructure programs for
local agriculture and economic growth. Finan et al (2015) conclude their review of the per-
sonnel economics of the state with the observation that when government bureaucracies
have to perform multiple complex tasks, questions of how to structure the principal-agent
relationships within these bureaucracies receive few answers from existing research. This
paper is beginning to ll the gap.
   We provide evidence consistent with economic theory of complex principal-agent re-
lationships in which non-pecuniary motivation and selection of di erent types of agents
in uence organizational productivity. We gather unique survey data to measure selection
traits–integrity, altruism, personality, public service motivation, risk aversion, cognitive
ability–of bureaucrats and politicians who make-up the complex organization of local gov-
ernment to provide the rst evidence of whether (which) selection traits are correlated with
organizational productivity. We provide robust evidence that it is political integrity in par-
ticular that matters for service delivery, to the exclusion of other traits. This evidence is
particularly relevant for understanding local state capacity for policy implementation in
poor countries. This is because budget and human resource constraints in poor countries
can limit the extent to which highly trained professionals can be attracted to the public
sector, or be given su cient incentives to make them accountable for their performance.
On the other hand, politics is ourishing in poor countries– with widespread participation
of citizens as voters and as contenders for political leadership at local levels (World Bank,
2016a)–whether for ill or good. The evidence in this paper shows that local politicians can
exert signi cant in uence over the functioning of bureaucracies at the frontlines. Policy
e orts to build capacity of the local state would do well to take seriously the role of political
selection, going beyond incentives, when crafting the principal-agent relationships (who
is tasked with doing what) in local bureaucracies.




                                               35
8    References

Abadie, A. and Kasy, M. 2017. The Risk of Machine Learning. Working Paper, MIT.
    Acemoglu, D., 2003. Why not a political Coase theorem? Social con ict, commitment,
and politics. Journal of Comparative Economics, 31: 620-652
    Acemoglu, D., and Jackson, M. 2015. History, Expectations, and Leadership in the Evo-
lution of Social Norms. Review of Economic Studies 82 (2): 423–56.
    Acemoglu, D., Kremer, M. and Mian, Atif Incentives in Markets, Firms, and Govern-
ments. The Journal of Law, Economics, and Organization, Volume 24, Issue 2, 1 October
2008, Pages 273–306, https://doi.org/10.1093/jleo/ewm055
    Ahmad E. and G. Brosio. 2006. Handbook of Fiscal Federalism, Edward Elgar.
    Ahmad E., and G. Brosio. 2009. Does Decentralization Enhance Service Delivery and
Poverty Reduction?. Studies in Fiscal Federalism and State-local Finance series. Edward
Elgar Publishing.
    Aidt, Toke, Golden, Miriam and Tiwari, Devesh, (2011), Incumbents and Criminals in
the Indian National Legislature, Cambridge Working Papers in Economics, Faculty of Eco-
nomics, University of Cambridge, https://EconPapers.repec.org/RePEc:cam:camdae:1157.
    Afridi, Farzana; Dhillon, Amrita; and Eilon Solan. 2017. Electoral Competition and
Corruption: Theory and Evidence from India. Working Paper.
    Aghion, Philippe; and Tirole, Jean (1997). The Journal of Political Economy, Vol. 105,
No. 1 (Feb., 1997), pp. 1-29.
    Akin, J.; Hutchinson, P. and Strumpf,K. 2005. Decentralisation and government pro-
vision of public goods: The public health sector in Uganda, The Journal of Development
Studies, 41:8, 1417-1443, DOI: 10.1080/00220380500187075
    Alesina, Alberto, and Eliana La Ferrara. 2005. Ethnic Diversity and Economic Perfor-
mance. Journal of Economic Literature 43: 721-61.
    Alesina, Alberto, Caterina Gennaioli, and Stefania Lovo. 2017. Public Goods and Ethnic
Diversity: Evidence from Deforestation in Indonesia. NBER WP 20504.
    Asher, Sam, and Paul Novosad. 2017. Politics and Local Economic Growth: Evidence


                                            36
from India. American Economic Journal: Applied Economics, 9 (1): 229-73.
   Ashraf, Nava, Oriana Bandiera, and Kelsey Jack. (2014). No Margin, No Mission? A
Field Experiment on Incentives for Public Services Delivery. Journal of Public Economics
120: 1-17.
   Ashraf, Nava, Oriana Bandiera, and Scott S. Lee. (2016), Do–gooders and Go–getters:
Selection and Performance in Public Service Delivery. Working Paper.
   Baber, Michael; Paul Kihn, and Andy Mo t. 2011. Deliverology: From idea to imple-
mentation. Washington, DC: McKinsey and Co.
   Bernheim, B.D. and Kartik, N., 2014. Candidates, Character, and Corruption. American
Economic Journal: Microeconomics, 6(2), pp.205-246.
   Besley, T., 2006. Principled agents? The political economy of good government. Oxford
University Press on Demand.
   Best, MC; Hjort, J. and David Szakonyi. 2017. Individuals and Organizations as Sources
of State E ectiveness, and Consequences for Policy. NBER Working Paper No. 23350.
   Besley, Timothy, and Maitreesh Ghatak. 2005. "Competition and Incentives with Moti-
vated Agents." American Economic Review, 95 (3): 616-636.
   Bidner, Chris, and Patrick Francois. 2013. “The Emergence of Political Accountability.”
Quarterly Journal of Economics 128 (3): 1397–448.
   Brosio, Giorgio (2000). Decentralization in Africa. Mimeo, International Monetary
Fund.
   Callen, Michael; Gulzar, Saad; Hasanain, Ali; Khan, Yasir; Rezaee, Arman. 2015. Person-
alities And Public Sector Performance: Evidence From A Health Experiment In Pakistan.
NBER Working Paper 21180. http://www.nber.org/papers/w21180.
   Cruz, Cesi; Keefer, Philip and Labonne, Julian. 2016. Incumbent advantage, voter in-
formation and vote buying. IDB Working Paper Series IDB-WP-711.
   Dal Bo, Ernesto; Finan, Frederico; Folke, Olle; Persson, Torsten; Rickne, Johanna. Who
Becomes a Politician? The Quarterly Journal of Economics, Volume 132, Issue 4, 1 Novem-
ber 2017, Pages 1877–1914, https://doi.org/10.1093/qje/qjx016
   Dal Bo, Ernesto; Finan, Frederico; Rossi, Martin. (2013). Strengthening State Capabil-

                                           37
ities: The Role of Financial Incentives in the Call to Public Service, Quarterly Journal of
Economics 128(3).
   Deserrano, Erika. (2017) Financial Incentives as Signals: Experimental Evidence from
the Recruitment of Village Promoters in Uganda. Working Paper.
   Dewatripont, Mathias, Jewitt, Ian and Tirole, Jean, (1999), The Economics of Career
Concerns, Part I: Comparing Information Structures, Review of Economic Studies, 66, issue
1, p. 183-198.
   Dixit, A. (2002). Incentives and Organizations in the Public Sector: An Interpretative
Review. Journal of Human Resources, 2002, vol. 37, issue 4, 696-727.
   Donato, Katherine, Grant Miller, Manoj Mohanan, Yulya Truskinovsky, and Marcos
Vera-Hernández. 2017. "Personality Traits and Performance Contracts: Evidence from a
Field Experiment among Maternity Care Providers in India." American Economic Review,
107 (5): 506-10.
   Elbers, Chris; Lanjouw, Jean O.; Lanjouw, Peter. (2002). Micro-level estimation of
welfare. Policy, Research working paper; no. WPS 2911. Washington, DC: World Bank.
http://documents. worldbank.org/curated/en/362131468739473297/Micro-level-estimation-
of-welfare.
   Ferraz, C. and Finan, F., 2008. Exposing Corrupt Politicians: The E ects of Brazil’s
Publicly Released Audits on Electoral Outcomes. The Quarterly Journal of Economics, ,
Vol. 123, No. 2: 703–745.
   Ferraz, C. and Finan, F., 2009. Motivating politicians: The impacts of monetary incen-
tives on quality and performance (No. w14906). National Bureau of Economic Research.
   Finan, Frederico; Olken, Benjamin; Pande, Rohini. 2015. The Personnel Economics of
the State. NBER Working Paper No. 21825. December.
   Finan, F. and Mazzocco, M. 2017. Electoral Incentives and the Allocation of Public
Funds. NBER Working Paper No. 21859
   Friedman, J.; Hastie, T.; and Rob Tibshirani. 2010. Regularization Paths for General-
ized Linear Models via Coordinate Descent. Journal of Statistical Software January 2010,
Volume 33, Issue 1.

                                            38
   George, Siddharth, Sarika Gupta, and Yusuf Neggers. 2018. Can We Text Bad Politicians
Out of O ce? Experimental Evidence from an Indian Election. Working Paper.
   Green, E., 2010. Patronage, district creation, and reform in Uganda. Studies in compar-
ative international development, 45(1), pp.83-103.
   Grossman, G. and Lewis, J.I., 2014. Administrative unit proliferation. American Politi-
cal Science Review, 108(1), pp.196-217.
   Grossman, Guy and Kristin Michelitch. 2018. Information Dissemination, Competitive
Pressure, and Politician Performance between Elections: A Field Experiment in Uganda.
American Political Science Review (forthcoming)
   Gulzar, Saad and Muhammad Yasir Khan. 2017. Motivating Political Candidacy and
Performance: Experimental Evidence from Pakistan. Working Paper.
   Jonas Hjort; Ethnic Divisions and Production in Firms , The Quarterly Journal of Eco-
nomics, Volume 129, Issue 4, 1 November 2014, Pages 1899–1946, https://doi.org/10.1093/qje/qju028
   John, O. P., Srivastava, S. (1999). The Big-Five trait taxonomy: History, measurement,
and theoretical perspectives. In L. A. Pervin; O. P. John (Eds.), Handbook of personality:
Theory and research (Vol. 2, pp. 102–138). New York: Guilford Press.
   Khemani, S.; Keefer, P. 2014. Mass Media and Public Education: The E ects of Access
to Community Radio in Benin. Journal of Development Economics, 109: 57-72.
   Khemani, S.; Keefer, P. 2015. The Government Response to Informed Citizens: New
Evidence on Media Access and the Distribution of Public Health Bene ts in Africa World
Bank Economic Review 2016 30:2, 233-267
   Laajaj, Rachid; Macours, Karen. 2017. Measuring skills in developing countries (En-
glish). Policy Research working paper; no. WPS 8000. Washington, D.C. : World Bank
Group. http://documents.worldbank.org/curated/en/775311488980295780/Measuring-skills-
in-developing-countries
   Moore, C., Detert, J.R., Trevino, L.K., Baker, V.L., and Mayer, D.M. (2012). Why employ-
ees do bad things: Moral disengagement and unethical organizational behavior. Personnel
Psychology, 65, 1-48.
   Muralidharan, Karthik, Paul Niehaus, and Sandip Sukhtankar. 2016. "Building State

                                            39
Capacity: Evidence from Biometric Smartcards in India." American Economic Review, 106
(10): 2895-2929.
       Olken,Benjamin A. (2007). Monitoring Corruption: Evidence from a Field Experiment
in Indonesia. Journal of Political Economy 115, no. 2 (April 2007): 200-249.
       Osborne, M., and Slivinski, A. (1996). A Model of Political Competition with Citizen-
Candidates. The Quarterly Journal of Economics, 111(1), 65-96. Retrieved from http://www.jstor.org/stable/2
       Perry, James L. (1996). Measuring Public Sector Motivation: An Assessment of Con-
struct Reliability and Validity. Journal of Public Administration Research and Theory: J-
PART, Vol.6, No 1.
       Pierre , Gaëlle, Maria Laura Sanchez Puerta, Alexandria Valerio, Tania Rajadel (2014).
STEP Skills Measurement Surveys: Innovative Tools for Assessing Skills. World Bank.
       Ra er, P., 2017. Does political oversight of the bureaucracy increase accountability?
Field experimental evidence from an electoral autocracy. Working paper, Harvard Univer-
sity.
       Rasul, I, D Rogger and M Williams (2017), “Management and bureaucratic e ectiveness:
A scienti c replication”, mimeo, UCL.
       Reinikka, Ritva; and Svensson, Jakob. 2004. Loca Capture: Evidence From a Central
Government Transfer Program in Uganda. The Quarterly Journal of Economics, vol. 119
(2).
       Rogger, Daniel (2017). Who Serves the Poor? Surveying Civil Servants in the Develop-
ing World. World Bank Policy Research Paper 8051.
       Tirole, Jean, (1994), The Internal Organization of Government, Oxford Economic Pa-
pers, 46, issue 1, p. 1-29, https://EconPapers.repec.org/RePEc:oup:oxecpp:v:46:y:1994:i:1:p:1-
29.
       Weaver, Je (2017). Jobs for Sale: Corruption and Misallocation in Hiring. Job Market
Paper, Yale University.
       World Bank. (2016a) Making Politics Work for Development: Harnessing Transparency
and Citizen Engagement. Policy Research Report. Washington, DC. © World Bank. https:
//openknowledge.worldbank.org/handle/10986/24461 License: CC BY 3.0 IGO.”

                                               40
   World Bank (2016b). Farms, cities and good fortune: assessing poverty reduction in
Uganda from 2006 to 2013. Washington, DC.




                                         41
Figure 1: Districts of Uganda - Survey Sample




                     42
                   Table 1: Politicians vs. Bureaucrats - Demographics


                                         Bureaucrat      Politician     Di      p-value       N
Individual characteristics
Female                                       0.18            0.34       0.16      0.00      2126
Age in years                                45.81           46.36       0.55      0.15      2126
Born in district where works                 0.39            0.78       0.38      0.00      2126
Household characteristics
Household members                            7.94           9.48        1.54      0.00      2125
Owns a car (percentage)                      0.64           0.46       -0.18      0.00      2122
Owns a TV (percentage)                       0.93           0.74       -0.18      0.00      2119
Owns a computer (percentage)                 0.74           0.49       -0.25      0.00      2120
Owns a DVD (percentage)                      0.84           0.68       -0.16      0.00      2119
Owns a fridge (percentage)                   0.69           0.47       -0.21      0.00      2119
Owns a gas stove (percentage)                0.39           0.24       -0.15      0.00      2118
Asset index - total                          2.27           1.66       -0.61      0.00      2123

Individual education
Less than Secondary                          0.00           0.21        0.20      0.00      2125
Less than College                            0.04           0.29        0.25      0.00      2125
Complete College                             0.47           0.40       -0.07      0.00      2125
Masters/PhD                                  0.49           0.11       -0.38      0.00      2125
Note: This table compares average characteristics of politicians and bureaucrats. Column (1) reports the
average characteristic for all politicians, while column (2) does the same for bureaucrats. Details about
the construction of variables are presented in the Annex. Column (3) presents the di erence between
politicians’ and bureaucrats’ means. Column (4) presents the p-value of a test with null hypothesis of
equal means. Column (5) reports the number of observations with non-missing values in each test.




                                                 43
                    Table 2: Politicians vs. Bureaucrats - Selection Traits


                                                          Bureaucrat     Politician    Di .    p-value        N
Panel A: Integrity measures
OK to spread rumors                                           0.05           0.11       0.06     0.00     2124
OK to borrow without permission                               0.02           0.04       0.02     0.01     2124
Not a sin to exaggerate credentials                           0.12           0.17       0.05     0.00     2124
Follow authorities orders                                     0.14           0.22       0.07     0.00     2124
No blame if group does the same                               0.04           0.09       0.05     0.00     2124
OK to take credit for others’ ideas                           0.20          0.23       0.02      0.21     2124
Some people lack feelings to be hurt                          0.11           0.13       0.02     0.18     2124
People are mistreated for a reason                            0.31           0.36       0.05     0.02     2124
Integrity Index                                               0.12          -0.20      -0.32     0.00     2124


Panel B: Non cognitive measures
Extraversion                                                   2.88          3.07       0.19     0.00     2126
Conscientiousness                                             3.34           3.24      -0.09     0.00     2126
Openness                                                      3.11           3.11      -0.01     0.75     2126
Emotional Stability                                            2.88          2.87      -0.01     0.72     2126
Agreeableness                                                 2.94          3.06       0.12      0.00     2126
Big ve index (mean of z-scores)                               -0.04          0.08       0.12     0.01     2126
Grit                                                          2.72           2.68      -0.04     0.07     2126
Decision making                                               3.26           3.13      -0.13     0.00     2126
Hostile Bias                                                  3.05           2.91      -0.14     0.00     2126
Other Non-Cognitive Index (Grit,Decision,Hostile)              0.12         -0.21      -0.34     0.00     2126
Non-Cognitive index (Big Five + Others)                       0.02          -0.04      -0.07     0.13     2126


Panel C: Public Sector Motivation
Attraction to Policy Making                                   3.51          3.51        0.00     0.83     2126
Commitment to Public Interest                                  4.08         4.07       -0.01     0.55     2126
Social Justice                                                4.09          4.08       -0.00     0.80     2126
Civic Duty                                                    4.05          4.10        0.05     0.00     2126
Compassion                                                    3.75          3.67       -0.07     0.00     2126
Self-Sacri ce                                                 4.05          4.14        0.08     0.00     2126
PSM Index                                                     -0.01         0.02        0.04     0.43     2126


Panel D: Cognitive and risk pro le measures
Altruism - percentage shared                                  0.29          0.32        0.04     0.00     2125
Choose dominated risk option                                  0.10          0.13        0.03     0.04     2126
Risk-aversion (1-5)                                           3.72          3.91        0.19     0.01     1869
Cognitive ability - Number of digits remembered               5.52          5.25       -0.28     0.00     2126

  Note: This table compares average characteristics of politicians and bureaucrats. Column (1) reports the
  average characteristic for all politicians, while column (2) does the same for bureaucrats. Details about
  the construction of variables are presented in the Annex. Column (3) presents the di erence between
  politicians’ and bureaucrats’ means. Column (4) presents the p-value of a test with null hypothesis of
  equal means. Column (5) reports the number of observations with non-missing values in each test.



                                                   44
                            Table 3: Selection Traits - Correlations

                  Education      Cognitive     Big Five    Non-Cognitive       Integrity    PSM      Altruism
Education               1
Cognitive            0.0380          1
Big Five             0.0339       0.0435*         1
Non-Cognitive        0.0736***    0.0776***     0.912***          1
Integrity            0.0451*      0.0512*       0.159***        0.218***          1
PSM                  0.0584**     0.0284        0.353***        0.385***        0.273***  1
Altruism            -0.0164       0.0615**      0.102***        0.102***       0.00573 0.0728***           1

        Note: All gures in this table refer to a pooled sample of bureaucrats and politicians.



                        Table 4: Descriptive Statistics - District Level


                                                                    Mean             SD             p50        N
Panel A - Health outcomes
DPT3 Coverage                                                       100.77          20.50          98.20       75
Deliveries in govt and PNFP facilities                               47.78          17.56          46.80       75
OPD Per Capita                                                       1.22           0.48           1.10        75
HIV testing in children born to HIV positive women                   93.26          83.69          71.60       75
Latrine coverage in households                                       75.21          16.18          78.20       75
IPT2                                                                 53.89          11.59          54.00       75
ANC4                                                                 34.69          10.92          33.80       75
TB success rate                                                      79.91          12.98          81.90       75
Approved posts that are lled                                         70.15          20.08          70.20       75
Monthly reports sent in on time (3)                                  88.87          11.64          92.40       75
Completeness monthly reports (2)                                     84.39          28.88         100.00       75
Completeness facility reporting (3)                                  95.99          7.03           99.50       75
Completeness of the annual report (2)                                70.85          34.19          84.80       75
Medicine orders submitted timely                                     94.00          10.22          96.40       75
Total Score Health DLT                                               73.30          7.25           75.00       75
Panel B - District socio-economic characteristics
District Population                                               324001.89      251864.32       251512.00 75
Urban population share                                              0.13           0.09            0.11    75
Poverty rate 2013                                                   0.23           0.16            0.23    75
Number of radios (HQ)                                               1.89           3.06            1.00    75
Voter Attachment to National Party - 2011                           0.68           0.15            0.70    75
Average Score UCE - 2011                                            53.37          3.31            54.04   75
 Note: This table presents descriptive statistics of health outcomes and characteristics of the 75 Districts
 included in the survey. Health outcomes are from the District League Tables produced by the Ministry
 of Health. The remaining district characteristics are from the 2010 Census and the Uganda Electoral
 Comission - details on the variables are provided in the Annex.



                                                   45
    Other Non-Cognitive


          Integrity index


            Asset Index


               Cognitive


 Choose dominated risk


                Altruism


    Non-Cognitive index


           Risk Aversion


             PSM index


          Big Five index

                            -.6   -.5   -.4   -.3   -.2   -.1   0     .1    .2    .3    .4     .5    .6
                                                     Point estimate and 95% CI



                    Figure 2: Politician vs. Bureaucrats Selection Traits

Note: This graph reports point estimates and 95 percent CI for the coe cient on "Politician" dummy
variable, in regressions with selection traits as dependent variable and district xed-e ects and individual
characteristics (gender, age education, household size and father education) as controls. Each reported
coe cient refers to a separate regression.




                                                    46
                                  1

                                             KS test p-value = 0.00

                                  .8
         Cumulative Probability




                                  .6



                                  .4



                                  .2



                                  0
                                       -4   -2                          0             2
                                                      Integrity Index

                                                 Politician             Bureaucrat



                                            Figure 3: Integrity

Note: The graph presents the cumulative distribution function for the Integrity index, separately for
politicians and bureaucrats. It also reports the p-value for the Kolmogorov-Smirnov (KS) test, under the
null hypothesis of equality of distribution.




                                                         47
                         1                                                                                                  1

                                                     KS test p-value = 0.29                                                             KS test p-value = 0.01

                         .8                                                                                                 .8
Cumulative Probability




                                                                                                   Cumulative Probability
                         .6                                                                                                 .6



                         .4                                                                                                 .4



                         .2                                                                                                 .2



                         0                                                                                                  0
                              -6     -4             -2               0                2   4                                      -4    -2                   0                2    4
                                                         PSM Index                                                                          Big five index (mean of z-scores)

                                                Politician               Bureaucrat                                                             Politician           Bureaucrat



                                    (a) Public Sector Motivation index                                                                 (b) Big Five Personality index



                         1                                                                                                  1

                                          KS test p-value = 0.00                                                                        KS test p-value = 0.32

                         .8                                                                                                 .8
Cumulative Probability




                                                                                                   Cumulative Probability




                         .6                                                                                                 .6



                         .4                                                                                                 .4



                         .2                                                                                                 .2



                         0                                                                                                  0
                              -4        -2                 0                  2           4                                      -4    -2                0                  2     4
                                    Other Non-Cognitive Index (Grit,Decision,Hostile)                                                  Non-Cognitive index (Big Five + Others)

                                                Politician               Bureaucrat                                                             Politician           Bureaucrat



                                   (c) Other Non-Cognitive Traits Index                                                               (d) Overall Non-Cognitive index



                                                          Figure 4: Distribution of Selection Traits - CDF

                         Note: The graphs present cumulative distribution functions for selection traits, separately for politicians
                         and bureaucrats. It also reports the p-value for the Kolmogorov-Smirnov (KS) test, under the null hy-
                         pothesis of equality of distribution.




                                                                                              48
    .4




                                                                                                           .5
                                                                                                           .4
    .3




                                                                                                              .3
Fraction




                                                                                                      Fraction
   .2




                                                                                                    .2
    .1




                                                                                                           .1
    0




                                                                                                           0
           0       .2                   .4            .6                         .8        1                                                Low risk aversion
                                                                                                                       Very low risk aversion                Medium risk aversion                  Very high risk aversion
                                                                                                                                                                                  High risk aversion
                                 Altruism - percentage shared                                                                                                 Risk-aversion (1-5)

                                    Bureaucrat                    Politician                                                                                                                                 Bureaucrat                                                      Politician



                                               (a) Altruism                                                                                                                                                               (b) Risk Aversion
    .4




                                                                                                           .5
                                                                                                           .4
    .3




                                                                                                            .3
                                                                                                     Fraction
Fraction




                                                                                                    .2
   .2




                                                                                                           .1
    .1




                                                                                                           0




                                                                                                                                             ary                                                  ary                                           lle
                                                                                                                                                                                                                                                      ge
                                                                                                                                                                                                                                                                                 lle
                                                                                                                                                                                                                                                                                       ge              hD
                                                                                                                                         prim                                                nd                                            co                               co                     's/P
                                                                                                                                                                                        co                                                                                                      ter
                                                                                                                                    te                                             se                                               lete                             lete                      s
                                                                                                                              ple                                           lete                                               mp                               mp                          Ma
                                                                                                                        c   om                                         mp                                                 co                               Co
                                                                                                                     an                                            o                                                  n
    0




                                                                                                                th                                              nc                                                tha
                                                                                                           ss                                             tha                                           Le
                                                                                                                                                                                                             ss
                                                                                                      Le                                             ss
               2        4                   6                8              10        12                                                        Le
                            Cognitive ability - Number of digits remembered                                                                                                                   Coarser measure of individual's education

                                    Bureaucrat                    Politician                                                                                                                                 Bureaucrat                                                      Politician



                                     (c) Cognitive Ability                                                                                                                                                          (d) Education level



                                                       Figure 5: Distribution of Selection Traits




                                                                                               49
2




                                                                                                    2
1




                                                                                                    1
0




                                                                                                    0
-1




                                                                                                    -1
-2




                                                                                                    -2
-3




                                                                                                    -3
          Slope = 0.29 [s.d. = 0.11]                                                                          Slope = 0.09 [s.d. = 0.11]
     -3                 -2                 -1             0                1           2                 -3                -2                -1              0                 1   2
                                  Integrity Index - Average Politicians                                                              Risk Aversion - Average Politicians



                                                 (a) Integrity Index                                                                                (b) Risk Aversion
2




                                                                                                    2
1




                                                                                                    1
0




                                                                                                    0
-1




                                                                                                    -1
-2




                                                                                                    -2
-3




      Slope = 0.07 [s.d. = 0.11]                                                                    -3    Slope = 0.00 [s.d. = 0.11]
          -2                 -1               0             1                  2           3             -1                  0                  1                   2              3
                                  Cognitive index - Average Politicians                                                  Choose dominated risk option - Average Politicians



                                                (c) Cognitive Index                                                                       (d) Dominated Risk Option
2




                                                                                                    2
1




                                                                                                    1
0




                                                                                                    0
-1




                                                                                                    -1
-2




                                                                                                    -2
-3




                                                                                                    -3




           Slope = 0.19 [s.d. = 0.11]                                                                     Slope = 0.08 [s.d. = 0.11]
     -2                               0                          2                         4                    -2               -1            0             1                 2   3
                                     Altruism - Average Politicians                                                             Non-Cognitive Index - Average Politicians



                                                        (e) Altruism                                                                 (f) Overall Non-Cognitive Index
2




                                                                                                    2
1




                                                                                                    1
0




                                                                                                    0
-1




                                                                                                    -1
-2




                                                                                                    -2
-3




                                                                                                    -3




               Slope = 0.08 [s.d. = 0.11]                                                                             Slope = 0.08 [s.d. = 0.11]
     -3                -2                -1             0              1           2                     -4                     -2                  0                      2       4
                                  Big Five Index - Average Politicians                                                                PSM Index - Average Politicians



                                     (g) Big Five Personality index                                                                                     (h) PSM index

                                                  Figure 6: DLT Score vs. Selection Traits - Politicians

     Note: The graphs present scatter plots, at the District Level, between the residual of a regression of DLT
     Score on District poverty rates, and selection traits of interviewees.


                                                                                               50
2




                                                                                          2
1




                                                                                          1
0




                                                                                          0
-1




                                                                                          -1
-2




                                                                                          -2
-3




                                                                                          -3
               Slope = -0.02 [s.d. = 0.11]                                                      Slope = 0.19 [s.d. = 0.11]
     -3              -2              -1            0             1       2                      -2                    -1                    0               1         2
                            Integrity Index - Average Bureaucrats                                                           Risk Aversion - Average Bureaucrats



                                               (a) Integrity Index                                                                         (b) Risk Aversion
2




                                                                                          2
1




                                                                                          1
0




                                                                                          0
-1




                                                                                          -1
-2




                                                                                          -2
-3




                  Slope = 0.11 [s.d. = 0.11]                                              -3    Slope = -0.06 [s.d. = 0.11]
     -4                    -2                  0                 2               4             -1                  0                1                 2               3
                            Cognitive index - Average Bureaucrats                                              Choose dominated risk option - Average Bureaucrats



                                             (c) Cognitive Index                                                                 (d) Dominated Risk Option
2




                                                                                          2
1




                                                                                          1
0




                                                                                          0
-1




                                                                                          -1
-2




                                                                                          -2
-3




                                                                                          -3




      Slope = 0.25 [s.d. = 0.11]                                                                          Slope = 0.06 [s.d. = 0.11]
     -2               -1              0             1                2       3                 -3                -2             -1             0             1            2
                               Altruism - Average Bureaucrats                                                         Non-Cognitive Index - Average Bureaucrats



                                                  (e) Altruism                                                              (f) Overall Non-Cognitive Index
2




                                                                                          2
1




                                                                                          1
0




                                                                                          0
-1




                                                                                          -1
-2




                                                                                          -2
-3




                                                                                          -3




      Slope = -0.01 [s.d. = 0.11]                                                               Slope = -0.23 [s.d. = 0.11]
          -2                 -1                0                1            2                       -2                -1              0             1            2           3
                            Big Five Index - Average Bureaucrats                                                             PSM Index - Average Bureaucrats



                                (g) Big Five Personality index                                                                                 (h) PSM index

                                             Figure 7: DLT Score vs. Selection Traits - Bureaucrats

     Note: The graphs present scatter plots, at the District Level, between the residual of a regression of DLT
     Score on District poverty rates, and selection traits of interviewees.


                                                                                     51
             Integrity index


                PSM Index


             Big Five Index


       Non-Cognitive Index


                   Altruism


     Dominated Risk Option


           Cognitive Ability


             Risk Aversion


                                      -.2              0                   .2        .4
                                                   Point estimate and 95% CI


                                             (a) Politicians’ Traits




              Integrity index


                 PSM Index


             Big Five Index


        Non-Cognitive Index


                    Altruism


     Dominated Risk Option


            Cognitive Ability


              Risk Aversion


                                -.4         -.2         0             .2        .4        .6
                                                    Point estimate and 95% CI


                                            (b) Bureaucrats’ Traits

                                Figure 8: DLT Score vs. Selection Traits

Note: These graphs report the point estimate and 95 percent CI for the coe cient on selection traits in
regressions using District level DLT Score as dependent variable. Each coe cient refers to a separate
regression with non-reported controls including regional dummies, log total population, poverty rate,
share of urban population, number of radio stations’ HQs, vote share of the presidential candidate of
the NRM, her ndahl index for councillors’ votes and average students’ score on UCE. CI are constructed
using robust standard errors.

                                                       52
                    Number Radios (HQ)


             Average Integrity - Politician


                    Altruism - Bureaucrat


               Risk Aversion - Bureaucrat


       Less than College education - Bur


                       Poverty Rate 2013


       Voter Attachment to National Party


                         UCE score 2011

                                              -.8   -.6     -.4     -.2   0   .2     .4    .6   .8
                                                             Point estimate and 95% CI


                                      (a) Bivariate Regressions



                    Number Radios (HQ)


             Average Integrity - Politician


               Risk Aversion - Bureaucrat


                    Altruism - Bureaucrat


                       Poverty Rate 2013


                         UCE score 2011


       Less than College education - Bur


       Voter Attachment to National Party

                                              -.6     -.4         -.2     0     .2        .4    .6
                                                             Point estimate and 95% CI


                                     (b) Multivariate Regression


      Figure 9: Robust Correlates of Health Service Delivery - LASSO selection

Note: This graph reports the point estimate and 95 percent CI for the coe cient on the variable in the axis,
in a bivariate regression using District level health outcomes as dependent variables. CI is constructed
using robust standard errors. The variables used here are those previously selected in a LASSO model.
All variables are standardized to have unit standard deviation.



                                                     53
                            Table 5: Correlates of Health Service Delivery

                                                           (1)         (2)         (3)         (4)         (5)
 Integrity Index - Average Politicians                     0.317***    0.231**    0.219       0.218        0.163
                                                          (0.108)     (0.104)    (0.137)     (0.137)      (0.145)
 Integrity Index - Average Bureaucrats                                                        0.024        0.232
                                                                                             (0.167)      (0.169)
 Poverty rate 2013                                                    −2.316     −1.958      −1.959      −0.664
                                                                      (1.642)    (1.263)     (1.279)      (0.954)
 Log Total Population                                                  0.057      0.223       0.209      −0.248
                                                                      (0.250)    (0.273)     (0.300)      (0.334)
 Urban population share                                               −1.157     −0.920      −0.940      −1.217
                                                                      (1.719)    (1.636)     (1.665)      (1.852)
 Average Score UCE - 2011                                             −0.062     −0.106**    −0.106**    −0.105**
                                                                      (0.042)    (0.047)     (0.047)      (0.049)
 Voter Attachment to National Party - 2011                            −1.459     −2.054      −2.082      −3.123*
                                                                      (1.098)    (1.409)     (1.443)      (1.545)
 Average Her ndahl index - Councilor                                   0.004      0.003       0.003        0.009
                                                                      (0.014)    (0.014)     (0.014)      (0.016)
 Number of radios (HQ)                                                 0.111**    0.117***    0.119***     0.139**
                                                                      (0.042)    (0.040)     (0.042)      (0.054)
 PSM Index - Average Politicians                                                  0.091       0.092        0.147
                                                                                 (0.144)     (0.146)      (0.163)
 Big Five Index - Average Politicians                                             0.222       0.240        0.579
                                                                                 (0.319)     (0.361)      (0.372)
 Non-Cognitive Index - Average Politicians                                       −0.327      −0.349      −0.684*
                                                                                 (0.311)     (0.360)      (0.371)
 Altruism - Average Politicians                                                   0.085       0.085        0.048
                                                                                 (0.146)     (0.147)      (0.138)
 Choose dominated risk option - Average Politicians                               0.030       0.025      −0.097
                                                                                 (0.138)     (0.140)      (0.155)
 Cognitive index - Average Politicians                                            0.069       0.079        0.104
                                                                                 (0.165)     (0.175)      (0.162)
 Risk Aversion - Average Politicians                                             −0.004      −0.002        0.051
                                                                                 (0.115)     (0.118)      (0.151)
 Asset Index - Average Politicians                                               −0.400**    −0.392**    −0.224
                                                                                 (0.179)     (0.188)      (0.171)
 PSM Index - Average Bureaucrats                                                                         −0.096
                                                                                                          (0.205)
 Big Five Index - Average Bureaucrats                                                                    −0.509
                                                                                                          (0.412)
 Non-Cognitive Index - Average Bureaucrats                                                                 0.499
                                                                                                          (0.383)
 Altruism - Average Bureaucrats                                                                            0.120
                                                                                                          (0.129)
 Choose dominated risk option - Average Bureaucrats                                                        0.050
                                                                                                          (0.120)
 Cognitive index - Average Bureaucrats                                                                   −0.096
                                                                                                          (0.180)
 Risk Aversion - Average Bureaucrats                                                                       0.168
                                                                                                          (0.159)
 Asset Index - Average Bureaucrats                                                                         0.182
                                                                                                          (0.375)
 Constant                                                 −0.000       3.935      6.461       6.651        3.216
                                                          (0.110)     (3.842)    (4.297)     (4.736)     (12.840)
 Observations                                             75          75         75          75           75
 R-Squared                                                 0.101       0.565      0.656       0.656        0.756
 P-value (Integrity Bureaucrat == Integrity Politician)                                       0.408        0.767

Note: This table reports regressions using District’s DLT Scores as dependent variables. Non-reported
controls also include Region dummies. Robust standard errors reported in parentheses (* p<0.1, ** p<0.05,
*** p <0.01)




                                                           54
        Completeness reports

                IPT2 coverage

         Reports sent on time

             Latrine Coverage

       Completeness facilities

     Antenatal care coverage

         Deliveries in facilities

                    Total Score

              Outpatients visit

               DPT3 coverage

 Completeness annual report

        Medicine Order timely

        TB treatment success

          HIV testing children

         Approved posts filled

                                    -.5                 0                     .5                    1
                                                      Point estimate and 95% CI



                    Figure 10: Politician Integrity and Health Outcomes

Note: This graph reports the point estimate and 95 percent CI for the coe cient on Average Politician
Integrity in regressions using District level DLT Score as dependent variable. Non-reported controls
include regional dummies, log total population, poverty rate, share of urban population, number of radio
stations’ HQs, vote share of the presidential candidate of the NRM, her ndahl index for councillors’ votes
and average students’ score on UCE. CI are constructed using robust standard errors. CI is constructed
using robust standard errors.




                                                 55
                 Table 6: Dependent variable - PCA of health components

                                                                  (1)       (2)        (3)        (4)       (5)
      Integrity Index - Average Politicians                      0.130     0.276**    0.352** 0.354**       0.292**
                                                                (0.111)   (0.106)    (0.152) (0.153)       (0.140)
      Integrity Index - Average Bureaucrats                                                  -0.109         0.121
                                                                                             (0.180)       (0.190)
      Poverty rate 2013                                                    1.487     1.977    1.985         3.373***
                                                                          (1.666)   (1.462)  (1.442)       (1.121)
      Log Total Population                                                 0.066     0.184    0.249        -0.225
                                                                          (0.271)   (0.263)  (0.282)       (0.354)
      Urban population share                                               0.406     0.791    0.883         0.686
                                                                          (1.510)   (1.518)  (1.556)       (1.850)
      Average Score UCE - 2011                                            -0.098** -0.104** -0.104**       -0.128*
                                                                          (0.038)   (0.049)  (0.046)       (0.071)
      Voter Attachment to National Party - 2011                           -0.738    -1.417   -1.292        -2.379
                                                                          (0.973)   (1.318)  (1.400)       (1.593)
      Average Her ndahl index - Councilor                                  0.023*    0.023    0.021         0.022
                                                                          (0.012)   (0.015)  (0.014)       (0.018)
      Number of radios (HQ)                                                0.152*** 0.165*** 0.159***       0.176***
                                                                          (0.045)   (0.044)  (0.044)       (0.056)
      PSM Index - Average Politicians                                               -0.014   -0.018         0.136
                                                                                    (0.153)  (0.156)       (0.176)
      Big Five Index - Average Politicians                                           0.128    0.045         0.426
                                                                                    (0.308)  (0.337)       (0.312)
      Non-Cognitive Index - Average Politicians                                     -0.204   -0.106        -0.538
                                                                                    (0.315)  (0.350)       (0.354)
      Altruism - Average Politicians                                                 0.041    0.040        -0.049
                                                                                    (0.153)  (0.157)       (0.125)
      Choose dominated risk option - Average Politicians                            -0.123   -0.100        -0.161
                                                                                    (0.137)  (0.137)       (0.156)
      Cognitive index - Average Politicians                                          0.147    0.101         0.207
                                                                                    (0.122)  (0.138)       (0.128)
      Risk Aversion - Average Politicians                                            0.105    0.096         0.171
                                                                                    (0.103)  (0.102)       (0.146)
      Asset Index - Average Politicians                                             -0.271   -0.305*       -0.178
                                                                                    (0.162)  (0.179)       (0.186)
      PSM Index - Average Bureaucrats                                                                      -0.228
                                                                                                           (0.215)
      Big Five Index - Average Bureaucrats                                                                 -0.600
                                                                                                           (0.531)
      Non-Cognitive Index - Average Bureaucrats                                                             0.599
                                                                                                           (0.447)
      Altruism - Average Bureaucrats                                                                        0.025
                                                                                                           (0.099)
      Choose dominated risk option - Average Bureaucrats                                                    0.008
                                                                                                           (0.104)
      Cognitive index - Average Bureaucrats                                                                -0.248
                                                                                                           (0.171)
      Risk Aversion - Average Bureaucrats                                                                   0.210
                                                                                                           (0.195)
      Asset Index - Average Bureaucrats                                                                     0.044
                                                                                                           (0.309)
      Constant                                                  -0.000     3.404      3.682      2.816      9.928
                                                                (0.115)   (3.833)    (3.891)    (4.160)   (10.610)
      Observations                                               75        75         75          75         75
      R-Squared                                                 0.017     0.520      0.634      0.638      0.745
      P-value (Integrity Bureaucrat == Integrity Politician)                                    0.080      0.495

Note: This table reports regressions using the rst component of a PCA of health outcomes as depen-
dent variables. Non-reported controls also include Region dummies. Robust standard errors reported in
parentheses (* p<0.1, ** p<0.05, *** p <0.01)




                                                           56
               Table 7: Integrity and Health Outcomes - by type of politician

                                                  (1)          (2)          (3)          (4)          (5)
Integrity Index - Average Politicians             0.231**
                                                 (0.031)
Integrity Index - Average Councillors                          0.180                                  0.276**
                                                              (0.111)                                (0.025)
Integrity Index - Average Chairperson                                       0.068                   −0.017
                                                                           (0.523)                  (0.905)
Integrity Index - Average Chair Contenders                                               0.136        0.090
                                                                                        (0.196)      (0.454)
Observations                                     75           75           72           67           65
R-Squared                                         0.565        0.555        0.566        0.555        0.614

  Note: This table reports regressions using District level health outcomes as dependent variables. Non-
  reported controls include regional dummies, log total population, poverty rate, share of urban population,
  number of radio stations’ HQs, vote share of the presidential candidate of the NRM, her ndahl index for
  councillors’ votes and average students’ score on UCE. CI are constructed using robust standard errors.
                     * p < 0.1, ** p < 0.05, *** p < 0.01. Standard errors in parentheses.




                                                   57
                               Latrine Coverage
                        Antenatal care coverage
                           Deliveries in facilities
                                  IPT2 coverage
                                      Total Score
                                Outpatients visit
                           Reports sent on time
                          Completeness reports
                         Completeness facilities
                                 DPT3 coverage
                          Medicine Order timely
                     Completeness annual report
                          TB treatment success
                           Approved posts filled
                             HIV testing children

                                                       -.5           0                       .5                  1
                                                                    Point estimate and 95% CI



                                                      (a) Councillors’ Integrity

                          Completeness reports
                           Deliveries in facilities
                                Outpatients visit
                          Medicine Order timely
                        Antenatal care coverage
                         Completeness facilities
                           Reports sent on time
                                  IPT2 coverage
                                 DPT3 coverage
                                      Total Score
                          TB treatment success
                               Latrine Coverage
                             HIV testing children
                           Approved posts filled
                     Completeness annual report

                                                       -.4    -.2            0          .2             .4        .6
                                                                    Point estimate and 95% CI



                                                      (b) Chairperson Integrity

                          Completeness reports
                         Completeness facilities
                           Reports sent on time
                                      Total Score
                                Outpatients visit
                                  IPT2 coverage
                                 DPT3 coverage
                           Deliveries in facilities
                               Latrine Coverage
                        Antenatal care coverage
                             HIV testing children
                     Completeness annual report
                          Medicine Order timely
                          TB treatment success
                           Approved posts filled

                                                      -.4    -.2         0         .2             .4        .6
                                                                    Point estimate and 95% CI



                                              (c) Chair Contenders Integrity

              Figure 11: Health outcomes vs. Integrity by politician’s type

Note: These graphs report point estimates and 95 percent CI for the coe cient on Average Politician
Integrity in regressions using District level DLT Score as dependent variable. Non-reported controls
include regional dummies, log total population, poverty rate, share of urban population, number of radio
stations’ HQs, vote share of the presidential candidate of the NRM, her ndahl index for councillors’ votes
and average students’ score on UCE. CI are constructed using robust standard errors. CI is constructed
using robust standard errors.
                                                                    58
                Table 8: Robustness - Other determinants of productivity

                                                                   (1)         (2)        (3)        (4)         (5)

                Integrity Index - Average Politicians             0.240**     0.230**    0.231**    0.239**     0.216*
                                                                 (0.103)     (0.106)    (0.107)    (0.106)     (0.109)
                Management Perception Index - Avg. Bureaucrats    0.133                             0.141
                                                                 (0.119)                           (0.117)
                Ethnic Her ndahl (All)                                       0.0322
                                                                             (0.117)
                Ethnic Her ndahl (Bureaucrat)                                           0.0248     0.00377
                                                                                        (0.145)    (0.136)
                Ethnic Her ndahl (Politician)                                           0.00696    0.0431
                                                                                        (0.128)    (0.126)
                Integrity Index - Average CAO/RDC                                                              0.00433
                                                                                                               (0.0930)
                Integrity Index - Average Health Bur.                                                          0.0921
                                                                                                               (0.0890)
                Poverty rate 2013                                -2.045      -2.280     -2.298     -2.024      -2.251
                                                                 (1.450)     (1.581)    (1.611)    (1.470)     (1.583)
                Log Total Population                             0.0485      0.0503     0.0508     0.0459      0.0705
                                                                 (0.249)     (0.250)    (0.251)    (0.247)     (0.252)
                Urban population share                           -1.368      -1.160     -1.169     -1.386      -1.339
                                                                 (1.708)     (1.729)    (1.750)    (1.756)     (1.735)
                Average Score UCE - 2011                         -0.0615     -0.0617    -0.0620    -0.0618     -0.0699
                                                                 (0.0401)    (0.0428)   (0.0429)   (0.0411)    (0.0418)
                Voter Attachment to National Party - 2011        -1.473      -1.483     -1.502     -1.462      -1.610
                                                                 (1.095)     (1.107)    (1.168)    (1.160)     (1.123)
                Average Her ndahl index - Councilor              0.00499     0.00522    0.00526    0.00636     0.00563
                                                                 (0.0135)    (0.0147)   (0.0149)   (0.0146)    (0.0144)
                Number of radios (HQ)                             0.118***    0.111**    0.111**    0.119***    0.110**
                                                                 (0.0410)    (0.0428)   (0.0430)   (0.0421)    (0.0427)

                Observations                                       75          75         75         75          74
                R2                                                0.578       0.565      0.565      0.579       0.572



Note: This table presents regressions using districts’ health outcomes as dependent variable and indi-
vidual traits and districts’ characteristics as explanatory variables. Unreported controls include regional
dummies.




                                                                 59
                                            Means by Group                     Di erence in Means (t-test)
                                    (1)         (2)            (3)         (4)           (5)              (6)
                                                                         Chair v.   Councillor v. Councillor v.
                                  Chair      Contender     Councillor   Contender      Chair         Contender
                                 mean/sd      mean/sd       mean/sd       di .          di .             di .
Integrity                          -0.19        0.02         -0.29         -0.20          -0.10          -0.31***
                                   (1.07)      (1.03)        (1.03)
PSM                                 0.17        0.17         -0.05         -0.00          -0.22          -0.22***
                                   (1.26)      (0.99)        (0.92)
Big Five                            0.19        0.15          0.03          0.04          -0.16           -0.12
                                   (1.03)      (0.98)        (1.04)
Other Non-Cognitive Traits         -0.01       -0.10         -0.28          0.09         -0.27**          -0.18**
                                   (0.98)      (0.90)        (1.01)
Non-Cognitive                       0.13        0.06         -0.11          0.07         -0.24**          -0.17**
                                   (1.00)      (0.97)        (1.03)
Cognitive                           5.63        5.41          5.12          0.22         -0.51***        -0.29***
                                   (1.14)      (1.25)        (1.11)
Choose dominated risk option        0.10        0.13          0.13         -0.03           0.04            0.01
                                   (0.30)      (0.33)        (0.34)
Altruism                            0.34        0.32          0.32          0.01          -0.02           -0.00
                                   (0.22)      (0.23)        (0.20)
Risk-aversion (1-5)                 3.92        3.65          4.01          0.27           0.09           0.36**
                                   (1.44)      (1.54)        (1.36)
Asset index - total                 2.60        2.26          1.30        0.34***        -1.30***        -0.96***
                                   (0.76)      (0.96)        (1.02)
Less than Secondary                 0.02        0.04          0.30         -0.01         0.28***         0.26***
                                   (0.16)      (0.19)        (0.46)
Less than College                   0.16        0.24          0.33         -0.08         0.17***          0.09**
                                   (0.37)      (0.43)        (0.47)
Complete College                    0.51        0.50          0.34          0.01         -0.17***        -0.16***
                                   (0.50)      (0.50)        (0.47)
Masters/PhD                         0.30        0.23          0.03          0.08         -0.27***        -0.19***
                                   (0.46)      (0.42)        (0.18)
Observations                        82          181           504           263            586             685


                         Table 9: Selection Traits by Type of Politician

  Note: This table compares average characteristics of elected Chairperson and Mayors; contenders for
  Chairperson/Mayor; and elected Councillors . Column (1) reports the average characteristic for all elected
  Chairperson/Mayors; column (2) does the same for non-elected Chairperson/Mayors; and column (3)
  for elected Councillors. Columns (4) presents the di erence in average trait between Chairpersons and
  Contender; column (5) between Chairperson and Councillors; and column (6) between Councillors and
  Contenders. The starts in columns (4)-(6) represent level of p-values in a t-test of equality of means (*
  p<0.1), ** p<0.05, *** p<0.01)




                                                   60
                         Table 10: Correlates of Politicians’ Integrity

                                                   (1)           (2)         (3)             (4)
                                                   All       Chaiperson   Contenders      Councillors
 Voter Attachment to National Party - 2011        0.304         2.720*       -1.341          0.290
                                                 (0.498)       (1.468)       (1.218)        (0.584)
 Average Her ndahl index - Councilor             -0.017**      -0.047**      -0.002         -0.014*
                                                 (0.007)       (0.019)       (0.018)        (0.008)
 Her ndahl index - Chairperson                   -0.003         0.005         0.004         -0.005
                                                 (0.004)       (0.012)       (0.012)        (0.005)
 Number of radios (HQ)                            0.008        -0.106*        0.076*        -0.007
                                                 (0.020)       (0.056)       (0.041)        (0.025)
 Poverty rate 2013                               -0.638        -1.047         0.112         -0.723
                                                 (0.473)       (1.158)       (1.433)        (0.548)
 Log Total Population                            -0.016        -0.065        -0.054         -0.024
                                                 (0.096)       (0.287)       (0.294)        (0.110)
 Urban population share                          -1.614**       0.736        -2.489         -1.377*
                                                 (0.706)       (2.035)       (1.742)        (0.833)
 Average Score UCE - 2011                         0.010         0.062        -0.011          0.017
                                                 (0.020)       (0.049)       (0.050)        (0.024)
 District Age 10-19 years                        -0.203        -0.642         0.072         -0.250
                                                 (0.159)       (0.405)       (0.319)        (0.201)
 District Age <= 10 years                        -0.022        -1.064**       0.333          0.032
                                                 (0.135)       (0.408)       (0.297)        (0.168)
 Female                                          -0.038         1.271**       0.437         -0.028
                                                 (0.086)       (0.519)       (0.336)        (0.093)
 Age in years                                     0.012***      0.004         0.010          0.014***
                                                 (0.004)       (0.014)       (0.009)        (0.005)
 Household members                               -0.010        -0.003        -0.004         -0.021*
                                                 (0.008)       (0.017)       (0.015)        (0.012)
 Less than College                                0.041        -0.963         0.018          0.024
                                                 (0.107)       (0.713)       (0.430)        (0.117)
 Complete College                                 0.064        -1.259*        0.119          0.107
                                                 (0.119)       (0.634)       (0.389)        (0.134)
 Masters/PhD                                      0.329**      -0.970         0.366          0.166
                                                 (0.162)       (0.701)       (0.447)        (0.257)
 Asset index - total                              0.010         0.230        -0.040          0.006
                                                 (0.042)       (0.189)       (0.106)        (0.050)
 Constant                                        -0.020        -1.462         1.680         -0.320
                                                 (1.545)       (4.015)       (3.967)        (1.877)
 Observations                                      768           82            181            503
 R-Squared                                        0.142        0.510          0.173          0.152

Note: This table reports regressions using the Integrity Index as explanatory variable, with the sample
restricted to politicians. Column (1) includes all politicians, while columns (2) - (4) restrict the sample to
Chairpersons, non-elected Contenders and Councillors, respectively..



                                                   61
             Table 11: Correlates of Politicians’ Integrity - Selection Traits

                                                Std    Unadjusted     Lower     Upper     Adjusted
                                 Coe cient     error    p-value       95% CI    95% CI    p-value
Public Sector Motivation
Attraction to Policy Making          0.14       0.07       0.05        0.00      0.28       0.49
Commitment to Public Interest        0.70       0.10       0.00        0.49      0.90       0.00
Social Justice                       0.38       0.10       0.00        0.20      0.57       0.00
Civic Duty                           0.24       0.10       0.02        0.05      0.44       0.20
Compassion                           0.56       0.08       0.00        0.40      0.71       0.00
Self-Sacri ce                        0.46       0.08       0.00        0.30      0.62       0.00
PSM                                  0.30       0.04       0.00        0.22      0.37       0.00
Non-Cognitive
Extraversion                          0.31      0.08       0.00         0.16     0.46       0.00
Conscientiousness                     0.50      0.09       0.00         0.32     0.68       0.00
Openness                              0.17      0.08       0.03         0.02     0.32       0.35
Emotional Stability                  0.05       0.08       0.59        -0.12     0.21       1.00
Agreeableness                         0.08      0.07       0.31        -0.07     0.22       0.96
Big Five                              0.16      0.04       0.00         0.09     0.23       0.00
Grit                                  0.13      0.08       0.10        -0.03     0.28       0.70
Decision making                       0.46      0.08       0.00         0.30     0.62       0.00
Hostile Bias                          0.06      0.07       0.44        -0.09     0.20       0.99
Non-Cognitive                        0.19       0.04       0.00         0.12     0.27       0.00
Altruism                              0.10      0.18       0.58        -0.25     0.45       1.00
Choose dominated risk option         -0.05      0.11       0.62        -0.27     0.16       1.00
Cognitive                             0.01      0.03       0.73        -0.05     0.07       1.00
Risk-aversion (1-5)                   0.01      0.03       0.82        -0.05     0.06       1.00
Asset index - total                  0.06       0.03       0.07        -0.00     0.12       0.58
Female                               -0.12      0.08       0.15        -0.27     0.04       0.79
Age in years                          0.01      0.00       0.00         0.00     0.02       0.04
Held political o ce before           0.04       0.08       0.60        -0.12     0.20       1.00
Complete College                      0.03      0.08       0.75        -0.13     0.18       1.00
Masters/PhD                           0.35      0.11       0.00         0.13     0.57       0.03

Note: This table reports univariate regressions using politicians’ integrity as dependent variables.
Columns (1) through (5) report report the results of the bivariate regressions for each of the selection
traits as explanatory variables of integrity. Column (6) presents p-values adjusted for multiple hypothe-
sis testing (using Holland’s method)




                                                 62
             Table 12: Correlates of Politicians’ Integrity - Selection Traits

                                                                 (1)
                                                              Integrity
               Public Sector Motivation
               Attraction to Policy Making               0.009         (0.067)
               Commitment to Public Interest             0.339***      (0.127)
               Social Justice                            0.016         (0.096)
               Civic Duty                               −0.209*        (0.108)
               Compassion                                0.244***      (0.082)
               Self-Sacri ce                             0.149         (0.092)
               Non-Cognitive Measures
               Extraversion                              0.213***      (0.082)
               Conscientiousness                         0.241**       (0.097)
               Openness                                 −0.138         (0.090)
               Emotional Stability                      −0.050         (0.091)
               Agreeableness                            −0.037         (0.079)
               Grit                                     −0.028         (0.077)
               Decision making                           0.135         (0.092)
               Hostile Bias                              0.010         (0.069)
               Other Measures
               Altruism                                  0.100         (0.175)
               Choose dominated risk option             −0.028         (0.104)
               Cognitive                                −0.014         (0.034)
               Asset index - total                       0.002         (0.039)
               Female                                   −0.083         (0.081)
               Age in years                              0.008**       (0.004)
               Held political o ce before               −0.078         (0.078)
               Complete College                         −0.062         (0.085)
               Masters/PhD                               0.157         (0.130)
               Constant                                 −4.964***      (0.748)
               Observations                            768
               R-Squared                                 0.329
Note: This table reports regressions using politicians’ integrity as dependent variables. Non-reported
controls also include District dummies.
                   * p < 0.1, ** p < 0.05, *** p < 0.01. Standard errors in parentheses.




                                               63
9     Annexes

9.1    Survey Data - Construction of Individual Selection Variables

Public Sector Motivation (PSM) – In order to measure bureaucrats’ and politicians’ moti-
vation to serve in the public sector, we use Perry’s Public Sector Motivation Index (Perry,
1996). Annex I presents the version used in the survey in Uganda: it is composed of 36
questions measuring motivation in six di erent dimensions – Attraction to Policy Making,
Commitment to Public Interest, Social Justice, Civic Duty, Compassion and Self-Sacri ce.
As discussed by Perry (1996), each dimension captures distinct motivations to serve iden-
ti ed previously in the public administration literature. Interviewees are asked how much
they agree with each of the questions on a scale of 1 (Strongly Disagree) to 5 (Strongly
Agree). We follow the established methodology in the personnel economics literature, as
in Dal Bo, Finan and Rossi (2013) and Callen et al. (2016): we rst recode all questions so
that higher values are associated with normatively positive answers (i.e. higher agreement
with “I think there are many public causes which are worth defending” but disagreement
with “I hardly think of the welfare of people I do not know personally”), and then create
an index for each dimension by taking simple averages of its answers. Finally, we create
an aggregate PSM Index for each individual by taking an average of the z-scores of each of
the six dimensions.


    Behavioral and Personality Traits – The module used to assess behavioral and person-
ality traits is the same as the one used in the World Bank STEP (Skills Towards Employa-
bility and Productivity) surveys . Annex II presents the 25 questions used to measure eight
distinct dimensions of personality. Taken together, the ve rst dimensions – Openness,
Conscientiousness, Extraversion, Agreeableness and Emotional Stability – are usually re-
ferred to as the Big Five Inventory (BFI), which have been shown to approximate the main
dimensions used to describe personalities (John and Srivastava, 1999). Whereas the BFI
concerns very broad personality traits, the module also collects data on three other more
speci c dimensions of behavior – Grit, Decision-Making and Hostility Bias. Interviewees

                                            64
are asked to rate each question from 1 (Almost never) to 4 (Almost always). Aggregation is
similar to the PSM methodology described above: we code variables such that higher values
indicate positive traits and obtain dimension-speci c indices by averaging over questions.
The Big Five Index is obtained as the average of z-scores in each of the ve dimensions,
whereas the Non-Cognitive Index is an average of the z-scores for the eight dimensions in
the module.


   Integrity – We use the Moral Disengagement Measure, developed by Moore et al. (2012),
to measure integrity values. Taken together, the eight questions presented in Annex III
have been shown to be correlated with unethical behavior in lab experiments. Individuals
are asked to rate their agreement with the statement on a scale of 1 (Strongly Disagree)
to 5 (Strongly Agree). We recode the variables such that higher values are disagreement
with unethical behavior, then proceed to take a simple average of the eight questions and
normalize it to obtain the Integrity Index.


   Risk Aversion - Risk preference is one of the fundamental parameters to understand in-
dividuals’ dynamic decisions (savings, consumption, investments in human capital, etc. . . ).
We measure risk preferences by using a common risk game: individuals are given the pos-
sibility of choosing, in a hypothetical game, between a certain gain of UGX 1 million and
a gamble with 50/50 probability with expected value higher than UGX 1 million. If the
individual chooses the certain prize the game ends; if they choose the gamble, they are
given another bet between the same certain prize and a lottery with lower expected value.
The game can last up to 4 rounds, giving us ve levels of risk-aversion (See Annex IV for
details). In order to screen if individuals understand the game, an initial question is asked
giving the interviewee the decision between a certain prize of UGX 1 million and a 50/50
bet between UGX 1 million and UGX 2 million. Choosing the previous is clearly dominated
by the latter, so individuals who choose that even after given a chance to reconsider are
marked as choosing a dominated risk option. We follow Dal Bo, Finan and Rossi (2013) in
using that choice as another measure of cognitive ability.

                                              65
       Cognitive Ability – The survey uses a digit span test to proxy for cognitive / memory
ability. We use a similar module to Weaver (2016), I which the interviewer reads out a string
of digits, starting with three, and the respondent has two chances to repeat them right. If
they can’t, the test ends; if they do remember, they are given a longer string; the game goes
on until the longest string with 12 digits. We code how many digits the individual could
remember, given two chances. As discussed above, we also use mistakes in choosing the
dominated option in the risk game as a measure of cognitive ability.


       Altruism – We measure altruism by the percentage of money shared in a hypothetical
game in which the individual is given UGX 50,000 and has to decide how much to keep
for themselves and how much to share with an anonymous person who received no money.




9.2       Reliability of selection traits

Here we brie y discuss the reliability of our three composite indices of selection traits:
Public Sector Motivation, Integrity and Non-Cognitive ability. Below we report the num-
ber of items, Cronbach’s alpha37 and the average inter-item covariance for each one of our
composite indices. For Public Sector motivation, we report results for each of the six indi-
vidual dimensions and for the overall PSM index; for Non-Cognitive ability, we separately
report results for each of the Big Five Index dimensions and the overall Big Five Index; as
well as for the remaining three dimensions of Grit, Decision-Making and Hostile Bias, and
the overall Non-Cognitive Index comprised of the eight dimensions above.
       For Public Sector Motivation, Cronbach’s alpha for each component varies from 0.2 to
  37
       Cronbach’s alpha is a measure of reliability used for psychometric tests. It is formally de ned as
                                                           ¯
                                                          Kc
                                               α=
                                                                ¯
                                                     ¯ + (K − 1)c
                                                     v

                                                                         ¯ is the average covariance between
                                  ¯ is the average variance of items and v
, where K is the number of items; c
all components. Cronbach’s alpha ranges from zero to one, and by construction if tends to one as the number
of items increase.



                                                       66
0.6, and average inter-item covariance from 0.5 to 0.12. The overall PSM index, however,
shows higher Cronbach’s alpha of 0.76.
   The Integrity index, comprised of eight items, has an alpha of 0.61 and average inter-
item covariance of 0.14.
   Each of the Big Five components show rather low alphas, between 0.2 and 0.4. These
results, however, are consistent with those found by Pierre et al. (2014) using the same
instrument in di erent low-income countries. It should be noted that we use a simpli ed
Big Five Instrument, with only 15 questions instead of the usual 44-question instrument.
The overall Big Five Index presents moderate reliability (alpha = 0.57). Finally, the overall
Non-Cognitive Index shows higher reliability with alpha = 0.68.
                           Table A1: Reliability of Selection Traits

                               Number of items          Cronbach’s alpha    Avg interitem covar.
 Attraction to Policy Making         4                       0.198                 0.056
 Commitment to Public Interest       7                       0.426                 0.057
 Social Justice                      5                       0.319                 0.049
 Civic Duty                          7                       0.549                 0.080
 Compassion                          7                       0.441                 0.089
 Self Sacri ce                       6                       0.613                 0.115
 PSM Index                          36                       0.765                 0.056
 Integrity Index                     8                       0.608                 0.143
 Extraversion                        3                       0.257                 0.060
 Conscientiousness                   3                       0.278                 0.044
 Openness                            3                       0.246                 0.053
 Emotional Stability                 3                       0.197                 0.036
 Agreeableness                       3                       0.437                 0.114
 Big Five                           15                       0.576                 0.042
 Grit                                3                       0.280                 0.067
 Decision Making                     4                       0.591                 0.120
 Hostile Bias                        2                       0.216                 0.070
 Non Cognitive                      24                       0.680                 0.042




9.3   District Level Administrative Data

Uganda Certi cate of Education (UCE) scores: we use data for 2011-2013 on average total
scores and division allocation of students at the end of lower secondary school. Students
take exams in six mandatory subjects and may choose four additional ones . Note that

                                              67
lower scores mean better performance (observed scores go from 8 to 72), and divisions are
broader categories of performance. In 2013, almost 300,000 students sat for the UCE exams.


                              District League Table (DLT): The District League Tables (DLT) are published in the An-
nual Health Sector Performance Report by the Ministry of Health and comprise both qual-
ity of care and management indicators. Data from local facilities is compiled using the
Health Management Information System (HMIS) and aggregated at the District level. We
use the 14 indicators for performance and the total score for the FY 2014/2015 (See below
for details).

 		                             		                                                  Weight	                 Explanation	
                                                                                              Diphtheria-tetanus-pertussis	(DTP3)	
                                DPT3	Coverage	(%)	                                    15	
                                                                                              immunization	coverage	
                                                                                              Share	of	infants	delivered	in	
                                Deliveries	in	gov’t	and	PNFP	facilities	(%)	          15	     government	or	Private	Non	for	Profit	
 Coverage	and	quality	care	




                                                                                              facilities	
                                                                                              Number	of	outpatients	department	
                                OPD	Per	Capita	                                       10	
                                                                                              (OPD)	visits	per	capita	
                                HIV	testing	in	children	born	to	HIV+	women	(%)	       10	     -	
                                Latrine	coverage	in	households	(%)	                   10	     -	
                                                                                              Share	of	pregnant	women	who	
                                IPT2	(%)	                                             5	      completed	Intermittent	preventive	
                                                                                              therapy	(IPT)	against	Malaria	
                                                                                              Proportion	of	pregnant	women	
                                ANC	4	(%)	                                            5	      receiving	4	or	more	antenatal	care	
                                                                                              visits		
                                TB	TSR	(%)	                                           5	      Tuberculosis	Treatment	Success	Rate	
                                Approved	posts	filled	(%)	                            10	     -	
                                %	Monthly	reports	sent	on	time	                       3	      -	
 Management	




                                %	Completeness	monthly	reports	                       2	      -	
                                %	Completeness	facility	reporting		                   3	      -	
                                Completeness	of	the	annual	report	(%)	                2	      -	
                                Medicine	orders	submitted	timely	(%)	                 5	      -	


                                            Figure 12: DLT Score - Components’ weights and description


                              2011 local and presidential elections: we use data from the 2011 elections, compiled by
the Uganda Electoral Commission. For Directly Elected Councilors, we compute all the
results at the electoral area level and then obtain the district-level average across all areas.
The Her ndahl Index is most commonly used as measure of rm competition in a given

                                                                               68
market, but has been widely used to measure political competition . The Her ndahl In-
dex is obtained, at the electoral area level, by the sum of the squares of candidates’ votes.
Higher values are associated with a less competitive environment (the index ranges from
1/N to 1, where N is the number of candidates).


   Poverty Rates: Poverty rates are estimates for 2013, using the national poverty line
(which ranges from US 0.88 to US 1.04 2005 PPP per capita depending on the region) (World
Bank, 2016b). These are small areas estimates produced by the World Bank, using data from
the 2013 UNHS and 2014 Census data. For details on methodology see for example Ebers,
Lanjow and Lanjow (2002).




                                             69
9.4   Questionnaire I - Integrity

  1. It is okay to spread rumors to defend those you care about.

  2. Taking something without the owner’s permission is okay as long as you’re just
      borrowing it.

  3. Considering the ways people often misrepresent themselves, it’s not a sin to exag-
      gerate your own credentials a bit.

  4. People shouldn’t be held accountable for doing questionable things when they are
      just doing what an authority gure told them to do.

  5. People can’t be blamed for doing things that are technically wrong if all their friends
      are doing it.

  6. Taking personal credit for ideas that were not your own is no big deal.

  7. Some people have to be treated roughly because they lack feelings that can be hurt.

  8. People who get mistreated have usually done something to bring it on themselves,
      or they are mistreated for a reason.




                                             70
9.5   Questionnaire II - Public Sector Motivation

  Attraction to Policy Making                    10. The duty of an o cer to the public
                                                     must always come before his loyalty
  1. Politics is a dirty word.
                                                     to his superiors
  2. I respect those public o cials who can
                                                 11. I consider public service my civic duty.
      turn a good idea into a law.
                                                     Social Justice
  3. Ethical behavior of public o cials is as
                                                 12. I think there are many public causes
      important as their skills.
                                                     which are worth defending
  4. The culture of “something for some-
                                                 13. I do not think that government can do
      thing” in public policy work does not
                                                     much to make society fair
      attract me.
                                                 14. If a group does not share in the pros-
      Commitment to Public Interest
                                                     perity of our society, then we all lose
  5. People can talk about the public inter-
                                                 15. I am willing to use every bit of my en-
      est, but what really concerns them is
                                                     ergy to make the world a more fair
      their own interest
                                                     place
  6. It is not easy for me to become
                                                 16. I am not afraid to defend the rights of
      strongly interested in what happens in
                                                     others, even if it means that I will be
      my community
                                                     mocked

  7. I contribute to my community in a               Civic Duty
      sel ess manner
                                                 17. When public o cials take the oath of
  8. Meaningful public service is very im-           public o ce, I think they accept obli-
      portant to me                                  gations not expected from other citi-
                                                     zens
  9. I would like to see public o cials do
      what is best for the community even        18. I am willing to go far in order to ful ll
      if this might hurt my interests                my obligations to my country

                                            71
19. Public service is one of the highest         29. I have little compassion for people in
    forms of citizenship                             need who are not willing to take the
                                                      rst step to help themselves
20. I think we all have a moral commit-
    ment to civic a airs, no matter how          30. There are not many public programs
    busy we are                                      that I wholeheartedly support

21. I have an obligation to care for the less        Self-Sacri ce
    wealthy
                                                 31. To make a di erence in society means
22. For me, the phrase “duty, honor and
                                                     more to me than personal achieve-
    country”, encourages deep emotions
                                                     ments
23. It is my responsibility to help solve
                                                 32. I believe in putting duty before self
    problems between people

    Compassion                                   33. Doing well      nancially is de nitely
                                                     more important to me than doing
24. I am not often touched by the plight of
                                                     good deeds
    the underprivileged

25. Most social programs are too vital to        34. To serve my fellow citizens would pro-

    eliminate                                        vide me with a good feeling even if no
                                                     one paid me for doing so
26. To me, patriotism includes concern for
    the welfare of others                        35. I think people should give back to so-

27. I hardly think of the welfare of people          ciety more than society gives to them

    I do not know personally
                                                 36. I’m one of those rare people who
28. Every day I am reminded of how often             would risk personal loss to help some-
    we depend on one another                         one else




                                            72
9.6   Questionnaire III - Behavior and Personality Traits

  Big Five Index                                10. Are you relaxed during stressful situ-
                                                    ations?
  Extraversion
                                                11. Do you tend to worry?
  1. Are you talkative?
                                                12. Do you get nervous easily?
  2. Do you like to keep your opinions to
                                                    Agreeableness
      yourself? That is, do you prefer to
      keep quiet when you have an opinion?      13. Do you forgive other people easily?

  3. Are you outgoing and sociable, for ex-     14. Are you very polite to other people?

      ample, do you make friends very eas-
                                                15. Are you generous to other people with
      ily?
                                                    your time or money?
      Conscientiousness

  4. When doing a task, are you very care-          Other Personality Traits
      ful?

  5. Do you prefer relaxation more than             Grit

      hard work?
                                                16. Do you nish whatever you begin?

  6. Do you work very well and quickly?
                                                17. Do you work very hard? For exam-
      Opennes                                       ple, do you keep working when others

  7. Do you come up with ideas other peo-           stop to take a break?

      ple haven’t thought of before?
                                                18. Do you enjoy working on things that

  8. Are you very interested in learning            take a very long time (at least several

      new things?                                   months) to complete?

  9. Do you enjoy beautiful things, like na-        Decision Making

      ture, art and music?                      19. Do you think about how the things
      Emotional Stability                           you do will a ect you in the future?

                                           73
20. Do you think carefully before you        you do will a ect others?
   make an important decision?
                                             Hostility Bias
21. Do you ask for help when you don’t
                                          23. Do people take advantage of you?
   understand something?

22. Do you think about how the things     24. Are people mean or not nice to you?




                                     74
9.7     Questionnaire IV - Risk Aversion Measurement

This annex describes the exact language used in the Risk Aversion module and how they
are used to construct both the Risk Aversion variable and the indicator that individual
chooses a dominated option in the risk game.


   I am now going to ask you a series of questions based on the following scenario: Imagine
you can choose between two bags. Once you have chosen one of the bags, you will put your
hand inside the bag and without looking you will pick a ball which will show the amount of
money you have won.


   Chooses dominated Risk Option: if individual chooses Bag No.1 in both questions
below, it’s agged as choosing dominated risk option.


   1. Now supposing Bag 1 has one ball that is worth UGX 1 million and Bag 2 has two balls:
one is worth UGX 1 million and the other ball is worth UGX 2 million. Which one of these two
bags do you choose; Bag 1 or Bag 2?
   2. Are you sure? You are going to pick only one ball from the bag you chose. Now if Bag
1 has two balls each worth UGX 1 million. If you choose Bag 1 you will win UGX 1 million.
If you choose Bag 2 you will win at least UGX 1 million or UGX 2 million depending on your
luck. Which one of these two bags do you choose; Bag 1 or Bag 2?


   Risk-Aversion


   The questions below are asked in sequence, whenever Bag 1 is chosen the game end.
Choices are coded as follow:


      • If Bag 1 is chosen in rst question: Very High Risk Aversion;

      • If Bag 1 is chosen in second question: High Risk Aversion;



                                             75
    • If Bag 1 is chosen in third question: Medium Risk Aversion;

    • If Bag 1 is chosen in fourth question: Low Risk Aversion;

    • If Bag 2 is chosen in fourth question: Very Low Risk Aversion;


   1. Now, imagine you can choose between the following two bags: Bag No. 1 guarantees
that you will win UGX 1 million. Bag No. 2 has a ball that is worth UGX 800,000 and another
ball that is worth UGX 2 million. Which one of these two bags do you choose; Bag 1 or Bag 2?


   2. And if now you could choose between: Bag No. 1, which again guarantees that you will
win UGX 1 million or Bag No. 2 which has a ball that is worth UGX 600,000 and another ball
that is worth UGX 2 million. Which one of these two bags do you choose; Bag 1 or Bag 2?


   3. Now, suppose you can choose between: Bag No. 1, which guarantees UGX 1 million or
Bag No. 2, which now has a ball that is worth UGX 400,000 and another ball that is worth
UGX 2 million. Which one of these two bags do you choose; Bag 1 or Bag 2?


   4. Now, suppose you can choose between: Bag No. 1, which guarantees UGX 1 million; or
Bag No. 2, which has a ball that is worth UGX 200,000 and another ball that is worth UGX 2
million. Which one of these two bags do you choose; Bag 1 or Bag 2?




                                            76
9.8   Surveyed Districts


  1. ADJUMANI       20. BUYENDE            39. KITGUM        58. NAMAYINGO

  2. AGAGO          21. DOKOLO             40. KOBOKO        59. NAMUTUMBA

  3. ALEBTONG       22. GOMBA              41. KWEEN         60. NEBBI

  4. AMOLATAR       23. GULU               42. KYANKWANZI    61. NGORA

  5. AMUDAT         24. HOIMA              43. LIRA
                                                             62. NWOYA
  6. APAC           25. IGANGA             44. LUUKA
                                                             63. OTUKE
  7. ARUA           26. ISINGIRO           45. LWENGO
                                                             64. OYAM
  8. BUDAKA         27. KAABONG            46. LYANTONDE
                                                             65. PADER
  9. BUDUDA         28. KABALE             47. MANAFWA
                                                             66. RAKAI
 10. BUGIRI         29. KABAROLE           48. MASAKA
                                                             67. RUBIRIZI
 11. BUIKWE         30. KABERAMAIDO        49. MAYUGE
                                                             68. RUKUNGIRI
 12. BUKEDEA        31. KALIRO             50. MBALE
                                                             69. SERERE
 13. BUKOMANSIMBI   32. KALUNGU            51. MBARARA
                                                             70. SHEEMA
 14. BULAMBULI      33. KAMULI             52. MITYANA
                                                             71. SOROTI
 15. BUNDIBUGYO     34. KAMWENGE           53. MOYO

 16. BUSHENYI       35. KANUNGU            54. MPIGI         72. SSEMBABULE

 17. BUSIA          36. KAPCHORWA          55. MUBENDE       73. WAKISO

 18. BUTALEJA       37. KASESE             56. MUKONO        74. YUMBE

 19. BUVUMA         38. KIRYANDONGO        57. NAKASONGOLA   75. ZOMBO




                                      77
10    Annex Tables

                     Table A2: Descriptive Statistics - Demographics




                                            Mean      SD      p25       p50      p75       N

     Individual characteristics
     Female                                  0.24     0.42    0.00     0.00      0.00    2126
     Age in years                           46.01     8.43    40.00    46.00    51.00    2126
     Born in district where works            0.53     0.50    0.00     1.00      1.00    2126
     Household characteristics
     Household members                       8.49     4.15    6.00     8.00     10.00    2125
     Owns a car (percentage)                 0.57     0.49    0.00     1.00      1.00    2122
     Owns a TV (percentage)                  0.86     0.35    1.00     1.00      1.00    2119
     Owns a computer (percentage)            0.65     0.48    0.00     1.00      1.00    2120
     Owns a DVD (percentage)                 0.78     0.41    1.00     1.00      1.00    2119
     Owns a fridge (percentage)              0.61     0.49    0.00     1.00      1.00    2119
     Owns a gas stove (percentage)           0.34     0.47    0.00     0.00      1.00    2118
     Asset index - total                     2.05     1.00    1.62     2.15      2.69    2123



     Individual education
     Less than Secondary                     0.08     0.27    0.00     0.00      0.00    2125
     Less than College                       0.13     0.33    0.00     0.00      0.00    2125
     Complete College                        0.44     0.50    0.00     0.00      1.00    2125
     Masters/PhD                             0.35     0.48    0.00     0.00      1.00    2125

        Note: All gures in this table refer to a pooled sample of bureaucrats and politicians.




                                                 78
                     Table A3: Descriptive Statistics - Selection Traits


                                                              Mean      SD      p25      p50     p75     N
Integrity measures
OK to spread rumors                                            0.07     0.26 0.00        0.00    0.00 2124
OK to borrow without permission                                0.03     0.17 0.00        0.00    0.00 2124
Not a sin to exaggerate credentials                            0.14     0.35 0.00        0.00    0.00 2124
Follow authorities orders                                      0.17     0.38 0.00        0.00    0.00 2124
No blame if group does the same                                0.06     0.24 0.00        0.00    0.00 2124
OK to take credit for others’ ideas                            0.21     0.41 0.00        0.00    0.00 2124
Some people lack feelings to be hurt                           0.12     0.32 0.00        0.00    0.00 2124
People are mistreated for a reason                             0.33     0.47 0.00        0.00    1.00 2124
Integrity Index                                                0.00     1.00 -0.61      -0.10    0.68 2124

Non cognitive measures
Extraversion                                                    2.95    0.48 2.67 3.00           3.33   2126
Conscientiousness                                               3.30    0.40 3.00 3.33           3.67   2126
Openness                                                        3.11    0.46 2.67 3.00           3.33   2126
Emotional Stability                                             2.87    0.43 2.67 3.00           3.00   2126
Agreeableness                                                   2.99    0.51 2.67 3.00           3.33   2126
Big ve index (mean of z-scores)                                -0.00    1.00 -0.69 -0.00         0.64   2126
Grit                                                           2.71     0.49 2.33 2.67           3.00   2126
Decision making                                                3.21     0.45 3.00 3.25           3.50   2126
Hostile Bias                                                    3.00    0.57 2.50 3.00           3.50   2126
Other Non-Cognitive Index (Grit,Decision,Hostile)              -0.00    1.00 -0.65 -0.00         0.69   2126
Non-Cognitive index (Big Five + Others)                        -0.00    1.00 -0.72 -0.01         0.68   2126

Cognitive and risk pro le measures
Altruism - percentage shared                                   0.30     0.22    0.10     0.30    0.40 2125
Choose dominated risk option                                   0.11     0.31    0.00     0.00    0.00 2126
Risk-aversion (1-5)                                            3.79     1.48    3.00     4.00    5.00 1869
Cognitive ability - Number of digits remembered                5.42     1.18    5.00     5.00    6.00 2126

Public Sector Motivation
Attraction to Policy Making                                    3.51     0.48 3.25        3.50    3.75   2126
Commitment to Public Interest                                  4.08     0.37 3.86        4.00    4.29   2126
Social Justice                                                 4.08     0.39 3.80        4.00    4.40   2126
Civic Duty                                                     4.07     0.38 3.86        4.00    4.29   2126
Compassion                                                     3.72     0.45 3.43        3.71    4.00   2126
Self-Sacri ce                                                  4.08     0.45 3.86        4.00    4.29   2126
PSM Index                                                      0.00     1.00 -0.64      -0.10    0.63   2126
        Note: All gures in this table refer to a pooled sample of bureaucrats and politicians.




                                                 79
                 Table A4: Selection Traits - Politicians vs. Bureaucrats

                     (1)           (2)          (3)              (4)                  (5)
                  Integrity       PSM        Big Five     Other Non-Cogn.        Non-Cognitive
Politician         -0.151** 0.239*** 0.242***                   -0.196***             0.0827
                  (0.0602) (0.0618) (0.0607)                   (0.0596)              (0.0610)
Constant           -0.719        -2.974*** -0.618               -1.434***             -1.094
                   (0.711)       (0.499)   (1.154)              (0.433)               (0.977)
Observations        2123         2125          2125             2125                   2125
R2                  0.137        0.108         0.093            0.129                  0.101
Note: This table reports regressions with selection traits as dependent variables and a dummy for politi-
cians as explanatory variable, using the entire sample of bureaucrats and politicians. Other controls
include gender, age, household size, whether individual was born in the same district where he works,
own and father’s education, and district dummies. The coe cients from this Table are presented in Figure
2




                 Table A5: Selection Traits - Politicians vs. Bureaucrats

                      (1)            (2)             (3)             (4)                (5)
                   Cognitive      Dominated       Altruism      Risk-Aversion       Asset index
Politician          -0.0401          0.0262         0.0388***         0.141           -0.0438
                    (0.0688)        (0.0194)       (0.0132)         (0.0922)          (0.0533)
Constant              3.997***      0.0856          0.261**          2.294**            0.234
                     (0.451)        (0.126)        (0.113)          (1.031)            (0.645)
Observations         2125            2125              2124          1869               2122
R2                   0.132           0.076             0.064         0.097              0.323
Note: This table reports regressions with selection traits as dependent variables and a dummy for politi-
cians as explanatory variable, using the entire sample of bureaucrats and politicians. Other controls
include gender, age, household size, whether individual was born in the same district where he works,
own and father’s education, and district dummies. The coe cients from this Table are presented in Figure
2




                                                 80
                                  Table A6: Di erences in pro le - Politicians

                                                       (1)           (2)     (3)       (4)       (5)       (6)       (7)       (8)
Integrity Index - Average Politicians                 0.231**
                                                     (0.104)
PSM Index - Average Politicians                                   0.112
                                                                 (0.104)
Big Five Index - Average Politicians                                       -0.030
                                                                           (0.095)
Non-Cognitive Index - Average Politicians                                            -0.056
                                                                                     (0.095)
Altruism - Average Politicians                                                                 -0.029
                                                                                               (0.125)
Choose dominated risk option - Average Politicians                                                        0.030
                                                                                                         (0.123)
Cognitive index - Average Politicians                                                                               0.077
                                                                                                                   (0.122)
Risk Aversion - Average Politicians                                                                                           0.120
                                                                                                                             (0.104)
Observations                                           75          75        75        75        75        75        75        75
R-Squared                                            0.565       0.546     0.538     0.539     0.537     0.537     0.540     0.546


  Note: This tables reports results from regressions having District DLT score as dependent variable and
  each of politicians’ average selection traits plus a series of controls as explanatory variables. Non-reported
  controls include regional dummies, poverty rate, log of total population, share of urban population, UCE
  scores in 2011, vote share for NRM in 2011, Her ndahl index for councillors’ votes in 2011 and number
  of radio HQ in District. The coe cients on this table are presented in Figure 8 .




                                                                81
                              Table A7: Di erences in pro le - Bureaucrats

                                                       (1)           (2)      (3)       (4)       (5)       (6)       (7)       (8)
Integrity Index - Average Bureaucrats                 0.008
                                                     (0.118)
PSM Index - Average Bureaucrats                                  0.057
                                                                (0.111)
Big Five Index - Average Bureaucrats                                        -0.042
                                                                            (0.120)
Non-Cognitive Index - Average Bureaucrats                                              0.037
                                                                                      (0.114)
Altruism - Average Bureaucrats                                                                   0.192*
                                                                                                (0.097)
Choose dominated risk option - Average Bureaucrats                                                        -0.019
                                                                                                          (0.118)
Cognitive index - Average Bureaucrats                                                                               -0.068
                                                                                                                    (0.145)
Risk Aversion - Average Bureaucrats                                                                                            0.176
                                                                                                                              (0.167)
Observations                                           75             75      75        75        75        75        75        75
R-Squared                                            0.537          0.539   0.538     0.538     0.564     0.537     0.539     0.554


  Note: This tables reports results from regressions having District DLT score as dependent variable and
  each of bureaucrats’ average selection traits plus a series of controls as explanatory variables. Non-
  reported controls include regional dummies, poverty rate, log of total population, share of urban popu-
  lation, UCE scores in 2011, vote share for NRM in 2011, Her ndahl index for councillors’ votes in 2011
  and number of radio HQ in District. The coe cients on this table are presented in Figure 8.




                                                               82
                                        Table A8: LASSO selection - regressions

                                                (1)          (2)         (3)         (4)        (5)       (6)        (7)         (8)         (9)

   Number Radios (HQ)                          0.478***                                                                                     0.302***
                                              (0.0864)                                                                                     (0.0803)
   Integrity - Politician                                  0.317***                                                                         0.238***
                                                           (0.108)                                                                         (0.0874)
   Risk Aversion - Bureaucrat                                          0.234*                                                               0.141
                                                                       (0.139)                                                             (0.0923)
   Altruism - Bureaucrat                                                           0.266**                                                  0.137**
                                                                                   (0.110)                                                 (0.0685)
   PSM - Politician


   Less than College education - Bureaucrat                                                   -0.312*                                      -0.211**
                                                                                              (0.160)                                      (0.0977)
   Poverty Rate 2013                                                                                    -0.316**                           -0.166
                                                                                                        (0.149)                            (0.122)
   Voter Attachment to National Party                                                                              -0.354***               -0.250**
                                                                                                                   (0.0969)                (0.110)
   Average UCE Score (2011)                                                                                                    -0.431***   -0.184*
                                                                                                                               (0.100)     (0.104)
   Constant                                   -0.296**    -1.34e-09   -5.46e-10   -1.98e-09   0.212     0.451**     1.559***   6.948***     4.262***
                                              (0.130)      (0.110)     (0.113)     (0.112)    (0.131)   (0.214)    (0.427)     (1.618)     (1.539)

   Observations                                 75           75          75          75         75        75         75          75          75
   R2                                          0.228       0.101       0.055       0.070      0.097     0.100       0.125      0.186        0.575



Note: This table presents regressions using districts’ health outcomes as dependent variable and individ-
ual traits and districts’ characteristics chosen by the LASSO method as explanatory variables. Columns
(1) - (9) present the results from bivariate regressions (equivalent to Figure 9a), while column (10) presents
the multivariate model (equivalent to Figure 9b)




                                                                         83
       ADJUMANI
  KIRYANDONGO
           OTUKE
         AMUDAT
          WAKISO
         BUVUMA
       KAABONG
           YUMBE
      BULAMBULI
 NAKASONGOLA
         BUDUDA
   BUNDIBUGYO
       BUSHENYI
           GOMBA
        ISINGIRO
       MANAFWA
       MBARARA
         RUBIRIZI
           ZOMBO
      RUKUNGIRI
       BUKEDEA
             GULU
           KALIRO
     KAMWENGE
         KASESE
         KOBOKO
            LUUKA
         MAYUGE
     NAMAYINGO
             NEBBI
         SHEEMA
          BUIKWE
             BUSIA
           MBALE
         SERERE
       BUTALEJA
       BUYENDE
      KABAROLE
  KABERAMAIDO
     LYANTONDE
        MITYANA
           NWOYA
          SOROTI
             ARUA
           BUGIRI
          KITGUM
        MUKONO
 BUKOMANSIMBI
          IGANGA
          KABALE
       KALUNGU
           KAMULI
    KAPCHORWA
        LWENGO
            PADER
   SSEMBABULE
           KWEEN
    KYANKWANZI
         MASAKA
             MOYO
             MPIGI
    NAMUTUMBA
       KANUNGU
      ALEBTONG
         BUDAKA
             RAKAI
           AGAGO
            HOIMA
               LIRA
           NGORA
         DOKOLO
       MUBENDE
             OYAM
      AMOLATAR
              APAC




                       -4                   -2                         0                         2
                                                     Integrity
                      excludes outside values


                 Figure 13: Distribution of Politicians’ Integrity within Districts

Note: This graph reports box-plots of politicians’ Integrity Index for each of the 75 Districts in the sample,
ordered from lowest to highest median Integrity. The vertical trade within the box represents the median
value while the box covers the 25th to 75th percentile. The right whiskers mark the smallest value above
Z, where Z is given by 75th percentile + 1.25 *IQR (interquartile range). The left whiskers is equivalent.




                                                   84