Pay and Mission: The Impacts of Performance Incentives in ...economics.virginia.edu/sites/economics.virginia.edu/files/public/Gine.pdf · of group-based performance incentives for

Pay and Mission: The Impacts of Performance Incentives in a

Multi-goal Development Organization

Xavier Gine, Ghazala Mansuri, and Slesh A. Shrestha∗

December 15, 2015

Abstract

Mission-oriented institutions with different organizational goals face a challenge when moti-vating employees. Workers of one such institution were randomly assigned to one of two bonusschemes, either incentivizing the performance of a microcredit program that contributed to theorganization’s sustainability, or incentivizing social mobilization aligned with the organization’smission of social and economic empowerment. The credit bonus was effective but only while itwas active and it undermined the mission. In contrast, the social bonus advanced the missionwithout compromising the microcredit program. For employees not assigned to teams, bothbonuses improved microcredit significantly, but they crowded out intrinsic motivation, withnegative impacts on performance of employees working in teams assigned to the social bonus.

∗Gine: Development Research Group, The World Bank, [email protected]. Mansuri: Development ResearchGroup, The World Bank, [email protected]. Shrestha: Department of Economics, National University ofSingapore, [email protected]. This project was jointly funded by the Development Research Group at the WorldBank, and the Pakistan Poverty Alleviation Fund. The views expressed herein are those of the authors and shouldnot be attributed to the World Bank, its executive directors, or the countries they represent.

1

Economic theory suggests that monetary incentives can be used to motivate workers to achieve

their employer’s goals.1 Tying pay to worker performance is also an increasingly popular tool to

improve the quality of public goods, and aid effectiveness (Montagu and Yamey, 2011; Olken et al.,

2014). Institutions that deliver such services include public bureaucracies and private nonprofits,

that have a broader mission; and they pursue this mission through diverse and multiple goals (Besley

and Ghatak, 2005). These mission-oriented organizations face the complex challenge of aligning their

workers’ interest with their overarching mission and goals.

Employees, in any organization, face trade-offs between their efforts towards different organi-

zational goals. For-profit firms overcome this multi-task problem through profit-based incentives

(Murphy, 1999). Mission-oriented organizations however, cannot easily define a single measure that

can be used to reward their employees’ effort. Incentives in this context can be ineffective or even

counterproductive when workers only focus on the incentivized tasks at the cost of the unrewarded

ones (Holmstrom and Milgrom, 1991; Baker, 1992). In addition, some organizational goals may

not easily translate into measurable indicators. Rewarding observed measures of performance that

do not truly reflect effort can weaken the worker’s response to such incentives. Workers in these

mission-oriented organizations may also be intrinsically motivated by the mission. Extrinsic rewards

through incentives can crowd out such motivation, and exacerbate the free-riding problem associated

with teamwork (Osterloh and Frey, 2000). In this complex setting, identifying incentives structures

that align worker and organizational incentives can facilitate the effective targeting of mission goals,

and therefore improve the effectiveness of these organizations.

In this paper, we study the effects of performance-based monetary incentives in mission-driven

organizations through a randomized pay-for-performance worker bonus scheme implemented by the

National Rural Support Program (NRSP) in Pakistan. NRSP is a nonprofit organization with core

missions of promoting social and economic empowerment, alongside enhancing the livelihoods of

poor households. In practice, this is achieved through the creation and strengthening of community

organizations (COs) and the provision of microcredit loans to CO members. NRSP’s twin goals are

carried out by its on-the-ground staff, also known as Field Assistants (or FAs), who are responsible

for delivering the services offered by NRSP to its clients.

We study the impacts of two types of bonus incentives for FAs. The social bonus rewards effort in

improving the quality of COs. The credit bonus on the other hand incentivizes the quality and scale

of NRSP’s microcredit program. Both incentives were designed to be based on objective measures

of performance. Adjustable targets were set according to past performance and bonuses, amounting

to at most 20 percent of the FA’s fixed monthly salary, were paid every month.

1See Predergast (1999) for a general overview of worker incentives used in organizations.

2

The experiment randomly assigned the 162 FAs who were working for NRSP across 15 districts

of Punjab and Sindh into one of the two bonus schemes or to the control group. We use independent

survey data along with administrative data from NRSP to answer the following set of questions: (1)

Do credit or social incentives elicit greater effort and improve performance on incentivized organiza-

tional goals? (2) Does the nature of the tasks being incentivized affect how they respond to extrinsic

rewards? (3) Are there trade-offs or complementarities between different organizational goals, and if

so, are incentives based on one goal better at achieving the organization’s overarching mission than

others? (4) Do monetary incentives crowd out intrinsic motivation, and what effects do they have

on teamwork and team performance?

We find that the introduction of the credit bonus improved the NRSP’s microcredit program but

only along the outcomes directly incentivized by the bonus. Moreover, any improvements in micro-

credit came at the costs of worsening the quality of COs managed by these FAs (Cr-FAs). The credit

bonus therefore undermined NRSP’s goal of empowering communities through CO mobilization. In

contrast, the social bonus significantly increased new CO formation, and it did so without worsening

the quality of the COs or the microcredit program. In fact, the social bonus was as effective as the

credit bonus at improving credit outcomes for employees not assigned to teams. This suggests that

incentivizing social tasks, at least when FAs are working individually, is likely to complement the

NRSP’s microcredit program, and it is more suited for aligning worker interests with the NRSP’s

overarching mission.

The positive impact of credit bonus on microcredit outcomes is explained by the increased amount

of time spent by Cr-FAs on microcredit activities. This effort however came at the cost of time that

was previously devoted to attending CO meetings. This reallocation of time results in the trade-offs

observed among Cr-FAs. In addition, we find that credit bonus also influenced the type of clients that

Cr-FAs target. In particular, they targeted clients who were better off and thus more creditworthy

undermining the empowerment goal. This shift in client selection by Cr-FAs is consistent with the

observed improvement in microcredit outcomes. But it also highlights a greater tendency among

Cr-FAs to select CO members with microcredit potential rather than poorer individuals lacking

empowerment. In comparison, Soc-FAs do not appear to reallocate their time or client focus in

accordance with the incentive.

The positive impact of the credit bonus on microcredit is however temporary as it disappears

when the bonus is terminated. The social bonus, in contrast, continues to influence the microcredit

outcomes of COs managed by Soc-FAs. The strong relationship between the quality of a CO and

its future creditworthiness suggests that investing in the process of social mobilization could offer a

more effective and sustainable strategy for targeting NRSP’s twin goals of community empowerment

3

and increasing client outreach for its microcredit program.

Both types of monetary incentives also crowd out intrinsic motivation in a way that there is

no change in overall effort. This decline in intrinsic motivation may however have affected team

cohesion through a larger proclivity to free ride when FAs were unable to monitor each other’s effort

on the incentivized tasks. As a result, the social bonus had a detrimental effect on team performance,

and it decreased the willingness of Soc-FAs to work in teams. In contrast, the impact of the credit

bonus is not affected by whether Cr-FAs are assigned to teams. These results highlight yet another

challenge that mission-oriented organizations face when using extrinsic rewards to motivate workers.

Our results contribute to the growing literature on the use of performance-based incentives to

improve the delivery of public goods and services in both developing and developed countries (Lavy,

2002, 2009; Duflo et al., 2012; Olken et al., 2014; Imberman and Lovenheim, 2015). The empirical

literature on this topic has thus far focused largely on the effects of incentivizing teachers (Glewwe et

al., 2010; Muralidharan and Sundararaman, 2011), healthcare providers (Basinga et al., 2011), and

government officials (Bo et al., 2013; Olken et al., 2014), by comparing individual- and group-based

incentives to the fixed-rate salaries for these workers; and by estimating spillover effects of incentives

on the non-incentivized workers or tasks.2

Our contribution to this literature is two-fold. First, we explicitly examine the tradeoffs between

incentivized and non-incentivized tasks that provide a more nuanced understanding of the impact

of performance-based incentives in organizations with multiple goals that are often interrelated.

Second, we extend the study of performance-based incentives to a development organization that

offers credit services alongside community mobilization, livelihoods and empowerment, and asset

transfers to the ultra-poor households (Mansuri and Rao, 2013).

While our experiment is conducted with a microcredit organization, our results also relate to the

broader literature on the role of performance pay in organizations in general (Lazear, 2000; Paarsch

and Shearer, 2000; Shearer, 2004). Studies that contemporaneously vary worker’s incentive structure

within a single firm are rare (with an exception of Bandiera et al. (2007, 2013)). Therefore, our results

contribute to answering important questions related to designing employment contracts, particularly

in mission-oriented organizations that face multitask problems (Besley and Ghatak, 2005), and likely

employ intrinsically motivated workers (Osterloh and Frey, 2000; Bowles and Polanina-Reyes, 2012).

2Muralidharan and Sundararaman (2011) uses a large-scale experiment that randomized individual and groupincentives to teachers in government schools in India, and find that both types of incentives improved student testscores on the incentivized and non incentivized subjects, with no evidence of any adverse gaming of the incentiveprogram by the teachers. Glewwe et al. (2010) in Kenya find similar positive impacts on test scores in the short runbut these impacts disappeared after the incentive program ended. Goodman and Turner (2013) report no impactof group-based performance incentives for teachers in NY schools due to incentives to free ride. In health, Basingaet al. (2011) compare performance-based incentive with the traditional input-based funding in Rwanda, and reportthe greatest effect on those services that had the highest payment rates and needed the least effort from the serviceprovider.

4

The rest of the paper proceeds as follows. Section 1 describes the NRSP’s organizational goals

and its overall mission, and outlines the experiment design. Section 2 discusses the data, Section

3 describes the empirical strategy, and Section 4 reports the results of the experiment. Section 5

concludes.

1 Context and Experiment

This project is collaboration between the National Rural Support Program (NRSP) and the World

Bank. NRSP is a non-profit development organization operating in Pakistan since 1991. Its activ-

ities have so far covered more than 2.5 million households, with 550,000 current clients in all four

provinces, making it the largest rural support program in the country in terms of outreach, staff and

development activities.

NRSP’s main mission is to reduce poverty by empowering communities and investing in their

livelihoods through microcredit. The social mobilization efforts revolve around local groups known

as Community Organizations (COs). Each CO typically comprises of 15 individuals, who live close

to each other in the same village. Depending on the local norms, the CO members may be all men,

all women, or both. CO members meet regularly to save and identify and work on local development

issues through the creation of community co-financed and co-managed infrastructure projects, and

the implementation of skill and management training programs.

In addition, NRSP also provides microfinance services to two-thirds of members of the CO,

usually in the form of single and monthly installment loans with a maturity of six to 12 months for

the purchase of agricultural inputs, livestock, and investments in household enterprises. All loans

have joint liability at the CO level although new loans are issued even if some CO members are

overdue.3 CO meetings also serve as the main venue for the receipt and repayment of loans. NRSP

views its microcredit service, along with its social mobilization efforts, as important tools to improve

the livelihoods of communities where it operates.

NRSP’s dual focus on social mobilization and microcredit is reflected in its branch management

structure. In each branch or Field Unit (FU), a Credit Officer (CrO) is in charge of the microcredit

program, while a Social Organizer (SO) handles the inclusion and social mobilization aspects of the

program. Field Assistants (FAs), who are at the bottom of this hierarchy, report to the CrO for all

issues related to microcredit and to the SO related to social mobilization. FAs are the institution’s

front line staff, working individually or in two to three-person teams, to engage directly with local

3Borrowers are required to find two guarantors, who can be members of the same CO. FAs use guarantors to exertpeer pressure, rather than enforcing repayment from them. A new borrower starts with a maximum loan size of PKR10,000, which can increase in intervals of up to PKR 5,000 with each successful loan cycle.

5

communities and CO members on a daily basis. Teamwork among FAs is encouraged by NRSP

management, to ensure that services are delivered without interruption when an FA falls sick or

leaves NRSP. Most FAs work from a village branch (VB) office of an FU, and they tend to live close

to the communities they work with.4

FAs have a wide range of responsibilities across the NRSP’s two main objectives. They assist

SOs with CO formation and strengthening, which includes attending CO meetings, ensuring that

COs maintain adequate records of meetings, attendance and savings, and they gather and forward

requests from the CO for skill training. They also do the first screening of loan applications, including

an assessment of the creditworthiness of potential borrowers, which typically involve a visit to the

applicant’s home, and are charged with ensuring timely loan repayment, which could include visits

to the borrower’s household as needed.

At the time of this study in 2005, NRSP was rapidly expanding due to its new partnership with

the Pakistan Poverty Alleviation Fund (PPAF). PPAF provided large financial support through

both grants and loans to promote NRSP’s development goals. In particular, its microcredit program

was growing very fast. The health of this program was seen as critical to the growth and survival

of NRSP as an institution, since microcredit funds were loans to the organization from PPAF that

needed to be repaid. As a result, FAs were spending much of their time on new client outreach,

processing of loans, and monitoring of repayment. However, the impact of this increased focus

on loan disbursement and repayment was also seen as potentially undermining NRSP’s mission of

community empowerment.

In this context, we worked with NRSP to design and implement two types of pay-for-performance

incentives for FAs to test the potential trade-offs (or complementarities) between its more immediate,

short-term objective of a healthy microcredit program and its long-term goals of social and economic

empowerment through social mobilization.

The study was conducted in FUs located in all 15 districts across Sindh and Punjab, where

NRSP was active in March 2005. This provided us with a sample of 162 FAs (in 35 FUs), who

were working with NRSP at the time. These FAs were randomly divided into three groups (two

treatment groups, and one control group). The randomization was done at the FU-level, so that all

FAs in an FU belonged to the same treatment or control group.5 FAs in the control group continued

as before, i.e., with a flat monthly salary. FAs in the two treatment groups received one of the two

new bonus incentives.

The bonus was announced to FAs in the treatment FUs in March 2005. An FA in a treatment4Each FU has two to three VB offices, on average.5This was done at the request of NRSP. Since CrOs and SOs usually work at the FU level and supervise all FAs

within the FU, NRSP did not want any variations within a management team.

6

FU became eligible to receive a bonus starting on April 2005. The bonus intervention lasted for

15 months, and was terminated at the end of June 2006. FAs were not informed in advance about

whether and when the bonus would end.

1.1 Bonus intervention

We worked with NRSP to devise individual, performance-based monetary incentives for FAs. The

incentive was provided as a simple salary bonus payable each month conditional on an FA meeting

his monthly targets. If an FA earned a bonus in any month, the bonus pay was simply added to his

base salary. The monthly base salary of an FA was about PKR 3000 (USD 50.54) at the time.6 The

largest bonus an FA could earn in any month was PKR 600 (20 percent of the base salary). FAs in

control FUs received just the base salary.

FAs in the treatment group received one of two bonus incentives. The first incentivized perfor-

mance on disbursement and loan recovery (the credit bonus). The tasks tied to theses goals typically

involve visiting applicant’s home for loan processing, and visiting borrower’s household to collect

repayments. The second bonus incentivized performance on observable correlates of CO quality-

new CO formation, regular CO meeting, and saving by CO members (the social bonus). Because

these tasks are relatively more complex and difficult to achieve, the outcomes might be less reflective

of FA’s own effort.7

The incentive scheme was designed to be easily understood, transparent, and fair. Each bonus

had two triggers. The first trigger determined whether an FA was eligible to receive a bonus.

Conditional on FA satisfying the first trigger, the second trigger determined the bonus amount.

The targets for bonus triggers were set at the VB or FU level. The targets were set based on past

performance of VBs, and was set at a level that would always be higher than the current level of

performance. At the same time, targets were also set so as to be achievable for FAs.

The Appendix Table A.1 presents the triggers and other details about the bonus. Because social

triggers are noisier and more difficult to change relative to credit triggers, we assigned more FUs

to the social bonus and control group compared to the number of FUs in the credit bonus during

randomization. This increases the power to accurately estimate changes due to the introduction of

the social bonus. Appendix Table A.2 presents the details of the list of FUs in the study and their

bonus assignments and the timeline of the study is presented in Appendix Figure A.1.

6The exchange rate in March 2005 was USD 1 = PKR 59.36.7In addition, social mobilization outcomes like savings and attendance are less costly for CO members to renege

on compared to defaulting on loans.

7

2 Data

The data for the analysis is derived from multiple sources. FAs were interviewed between January

and February 2005 (prior to the announcement of the bonus). This baseline survey asked each

FA about his demographic and household characteristics, current employment conditions and work

history, along with measuring his effort, time allocations, and intrinsic motives for working with

NRSP. FAs who were still employed with NRSP at the end of the study period were interviewed

again in June 2006 (the last month of the bonus intervention), and asked questions that were similar

to the baseline interview.

These two rounds of survey data are supplemented with NRSP’s administrative data. We ob-

tained NRSP’s monthly employee records that include each FA’s employment status, salary and

bonus information, and the name of FU and VB where the FA worked. NRSP also provided us with

a monthly record of COs managed (or co-managed) by each FA from June 2004 to June 2006. This

FA-CO panel helps us identify each FA’s monthly portfolio of COs for the entire bonus intervention

period (15 months), and for 10 months before the bonus began.

The data on loan disbursements and recovery is obtained from NRSP’s Management Information

Systems (MIS) database. The MIS digitally collects financial data for each client (CO members) on

loans taken and repaid, by installment. The total of 5,364 unique COs appear in the MIS database

during the 25 months that overlap with the FA-CO panel. Out of them, 4,404 COs (82.1%) were

managed by 162 FAs who make up our study sample.8 4,008 COs (out of 4,404 COs) show loan

activity at least once during these 25 months, and have 5.81 active borrowers each month on average,

with a repayment or disbursement in 14 out of 25 months.

The information on social mobilization is obtained from the Monthly Progress Reports (MPRs)

submitted by each FA for all COs visited that month. The report includes information on meeting

attendance, member savings, and loans approved and denied during the meeting. The MPRs data

is available for 15 months when the bonus was implemented. These data was verified by a CrO

through random visits to a subset of scheduled CO meetings. Such visits were conducted in 15.4

percent of CO meetings held during the bonus period. We rely on the verified MPRs data rather

than the self-reported data for our analysis on the social outcomes.9

We aggregate the CO-level credit and social outcomes for each FA by month using information

8The rest of the COs were managed (and formed) by FAs who were hired after March 2004 (after the bonusintervention began). We do not include these new FAs and the COs that they manage in our analysis.

9We do not find any evidence of a different propensity of visit by a CrO across COs managed by FAs belongingto different treatments or control groups. In addition, selection into CrO visits as measured by the quality of the COvisited, was not different across treatment types of across the treatment and control groups. The quality measure ofthe CO was based on disbursements and repayments. A detailed discussion and analysis of a potential selection intoCrO visits is presented in Appendix B and Appendix Table B.1.

8

from the FA-CO panel. We calculate the performance of an FA before and after the bonus, by taking

the average across 9 months prior to the bonus announcement and the average over the 15 months

during the bonus period, respectively. The data from March 2005 (the month when the bonus was

announced but not yet implemented) is dropped from the analysis.10

Table 1, Columns 1, 2 and 3 report means of FAs’ baseline variables in the control, credit bonus,

and social bonus groups respectively. Columns 4, 5 and 6 report the P-values from the F-tests of

the difference in means between control FAs and credit bonus FAs (Cr-FAs); control FAs and social

bonus FAs (Soc-FAs); and Cr-FAs and Soc-FAs, respectively. Across all the reported variables, we

cannot reject that means are equal between the three groups at conventional levels of statistical

significance. As indicated by the F-test statistics at the bottom of Columns 4, 5 and 6, we also

cannot reject the joint equality of means across the full set of variables between control FAs and

Cr-FAs; control FAs and Soc-FAs; and Cr-FAs and Soc-FAs.

FAs in our study are on average 28 years old, roughly one-fourth are female, and slightly more

than half of them have at least a high school degree (equivalent to 12 years of education). The

average duration of employment with NRSP is 26 months, and more than 80 percent of them work

from a VB office. Across the sample, we find that FAs co-manage some of their CO-portfolio with

other FAs. Roughly one-third of FAs co-manage their entire CO-portfolio with other FAs, while

slightly less than one-fourth manage all their COs independently. Appendix Figure C.1 depicts

the distribution of the share of FA’s CO-portfolio that are co-managed with other FAs (over the 9

months before the bonus was announced). The median level of co-management is 73 percent, and we

categorize FAs who co-manage more than this median value as “partnered” FAs in the analysis.11

FAs on average manage 14 COs every month, and spend three-fourth of their time visiting house-

holds for loan disbursement and recovery (as compared to attending CO meetings). Their average

portfolio consists of 91.4 active loans (new and ongoing), with roughly PKR 100,000 disbursed each

month. The mean recovery rate on installments due at the end of each month was almost 99 percent,

while only 70 percent of such installments being fully recovered by the 20th of that month.

Slightly more than half of the FAs prefer the bonus to be paid on credit outcomes compared to

social outcomes. But 90 percent reports that social mobilization also helps achieve good repayment.

During the baseline interview, each FA was asked to rank the things that they like most about

working with NRSP. Roughly half of the FAs reported that the ability to help people is what they

like most. One-fifth of them also had done volunteer work before joining NRSP.

10FAs were informed in March 2005 that bonus would start in April. FA’s performance in March is likely to beaffected by this announcement. The results are however robust to including the month of March in the analysis.

11In the baseline, almost 90 percent of partnered FAs (71 out of 81 partnered FAs) co-manage their COs withone other FA, while the rest co-manage with two other FAs. We also found that FA partnership-groups were stablethroughout the study period (unless one of the partners quit NRSP).

9

Out of 162 FAs in our sample, 132 FAs were successfully interviewed in the follow up survey.

The attrition rate in the follow up survey is almost identical and not statistically different across

the three groups (see bottom of Table 1), suggesting that attrition bias is not a prominent concern

when examining impacts of bonus on outcomes measured in the follow up survey. In addition, CO

meetings held by 31 FAs were never visited by a CrO. These FAs do not appear in the verified

MPRS data, restricting our sample to 131 FAs when examining social outcomes. This selection rate

is however not statistically different across control FAs, Cr-FAs, and Soc-FAs (see bottom of Table

1).

Appendix Tables D.1 and D.2 present means of baseline variables in the control and the two bonus

groups, and their differences for the follow up and verified MPRS restricted samples, respectively.

Four out of 69 differences in means in Appendix Table D.1 and two differences in Appendix Table

D.2 are statistically significant at the 10 percent level, which is about what would be expected to

occur by chance. In both samples, the F-tests in Columns 4-6 do not reject the jointly equality of

means across the three groups at conventional levels of statistical significance.

3 Empirical Strategy

We estimate the impact of the bonus intervention by estimating OLS with the following specification:

Yi,1 = α + βr + θ Yi,0 + γ TCi + λ TSi + ε

where Yi,1 is the post-treatment outcome of interest for FA i, βr is a region dummy (one for each of

the four regions), and Yi,0 is the pre-treatment outcome for FA i. TCi is an indicator variable that

takes the value of one if FA i was assigned the credit bonus incentive, and zero otherwise. TSi takes

a value of one if FA i received the social bonus incentive, and zero otherwise, and ε is a mean-zero

error term. Standard errors are clustered at the FU level.

The coefficients of interest in the regression are γ and λ, which estimate the causal impact of the

credit and social bonus on FA outcomes Yi,1, respectively.

We also examine the impact of bonus separately for partnered and non-partnered FAs, as defined

in Table 1. The differential effect of bonus by FA’s baseline partnership is estimated using the

following specification:

Yi,1 = α + βr + θ Yi,0 + η Pi + γ TCi + λ TSi + δ Pi ∗ TCi + σ Pi ∗ TSi + ε

where Pi is an indicator variable that takes the value of one if FA i is a partnered FA, i.e. co-managed

10

more than 73 percent of his pre-treatment CO portfolio with other FAs.

The coefficients δ and σ on the interaction terms Pi ∗ TCi and Pi ∗ TSi respectively reveal the

extent to which the impact of credit bonus and social bonus varies according to whether an FA is

working independently or jointly in partnership-teams. The coefficients γ and λ estimate the impact

of credit and social bonus, respectively, on non-partnered FAs. The sums of the coefficients γ+δ and

λ+σ estimate their impacts on partnered FAs.

4 Results

4.1 Bonus payments and the nature of incentivized tasks

During the 15-months bonus period, 26.7 percent of treated FAs qualified for a bonus each month;

and 62.9 percent of those qualified on target A. The average bonus amount was PRK 570.30 for

target A and PKR 416.70 for target B. This indicates that bonuses were actually paid out every

month and that FAs took them seriously.

Comparing across the two types of bonus in Figure 1, two-fifths of Cr-FAs qualified for a bonus,

while only one-fifth of Soc-FAs qualified for a bonus in any given month. Soc-FAs received PKR

124.30 less on monthly bonus payment compared to Cr-FAs, who on average earned PKR 249.30 in

bonus each month. Soc-FAs therefore earned 50 percent less in bonus relative to Cr-FAs.

The higher rate of bonus incidence among Cr-FAs compared to Soc-FAs may reflect the greater

difficulties faced by FAs to affect CO quality relative to microcredit outcomes, especially during the

short period that the bonus was active. We find evidence consistent with this in the follow-up survey.

FAs were asked to rate how easy it was to improve various microcredit and CO quality outcomes.

Among control FAs, 56.8 percent claim that organizing a new CO is difficult and 43.2 percent that

improving attendance and encouraging members to save is hard. In contrast, only 18.9 percent find

expanding disbursement and ensuring good repayments difficult. The difference in the share of FAs

who find CO quality versus microcredit outcomes difficult to change is statistically significant at the

1 percent level.

This is consistent with the fact that social mobilization activities are relatively more complex

and therefore they may not correlate well with the FA’s own efforts on these tasks. Related, it may

be harder for FAs working in teams to assess the effort put by their fellow co-workers and as a result,

Soc-FAs might not respond as strongly to incentives and may free ride if assigned to a team.

11

4.2 Bonus impacts on NRSP’s twin goals

We examine the effects of the credit and social bonus on NRSP’s two organizational goals that were

incentivized by the two bonus schemes in Panel A of Tables 2 and 3, respectively.

4.2.1 Microcredit program

Table 2 examines the effects of the credit and social bonus on FA’s performance on microcredit

outcomes. Columns 1 and 2 present the impact on the two outcomes that triggered a payment of

the credit bonus: number of active loans and repayment by the 20th of the month. The number

of active loans increased by 24.1, and the repayment improved by 8.4 percentage points for Cr-FAs

(FAs offered the credit bonus incentive) compared to control FAs. Both estimates are statistically

significant at the 5 percent level. The size of these impacts is large, amounting to a 20 percent

increase in active loans and a 12 percent improvement in repayment over the mean performance

of control FAs. Cr-FAs also performed statistically significantly better than Soc-FAs (FAs with a

social bonus incentive) on repayment. The impacts of social bonus on the two trigger variables of

the credit bonus are small (9.406 and -0.020), and not different from zero at conventional levels of

statistical significance.

Columns 3-5 estimate the impacts of the bonuses on microcredit outcomes that were not directly

incentivized: number of new loans, disbursement amount, and repayment by end of the month.

In contrast to the impacts on the two trigger variables, Cr-FAs showed no improvements on any

of these non-incentivized microcredit outcomes. In fact, the improved repayment rate at the 20th

of the month made very little difference to the repayment rates at the end of the month, partly

because end of the month repayments were already higher than 96 percent among control FAs. The

impacts on new loans and disbursement amount are also small in magnitude (6.1 and 0.21 percent

respectively compared to the means in the control group), and are not statistically significantly

different from zero. It thus seems that there are little positive spillovers from the trigger outcomes

on non-incentivized outcomes. On these non-incentivized outcomes, the performance of Cr-FAs is

also not statistically significantly different from that of Soc-FAs. The impacts of social bonus on

new loans, disbursement amount, and end of the month repayment are 0.616, -444.8, and -0.003

respectively, and none of these estimates are statistically significant.

Since the outcomes in Table 2, Columns 1-5 may be correlated, we follow Kling et al. (2007) and

construct a summary index that aggregates information over multiple outcomes to account for this

problem of multiple inference. The microcredit index in Table 2, Column 6 is calculated by taking

an equally weighted average across the standard distributions of all five microcredit outcomes. The

impact of credit bonus on this index is positive and significant at the 10 percent level, suggesting

12

that Cr-FAs performed 0.238 standard deviations higher (on the microcredit index) than control

FAs. But this impact is largely due to improvements on the two specific outcomes that were directly

incentivized by the credit bonus, rather than a general increase in performance on all microcredit-

related tasks.

The impact of credit bonus on the microcredit index is also significantly different from that of

social bonus at the 10 percent level. The impact of social bonus is close to zero (0.002 standard

deviations) and not statistically significant. This suggests that the social bonus does not undermine

the performance of Soc-FA’s on microcredit outcomes.

4.2.2 CO quality

Table 3 estimates the effect on FA’s performance on various outcomes of CO quality. Columns 1-3

report the impacts on the three social bonus measures used as triggers, while Columns 4-6 report

the impacts on other CO quality measures not directly incentivized.

We find that Soc-FAs formed 0.225 more new COs each month than control FAs (Column 1).

The estimate is statistically significant at the 5 percent level, and amounts to a 58.6-percent increase

in CO formation compared to the mean in the control group (0.384). It is however almost identical

to and also not significantly different from Cr-FAs, who also increased CO formation by 0.284

compared to control FAs. The effect of credit bonus on CO formation is statistically significant at

the 10 percent level, and not entirely unexpected since CO membership is a prerequisite for applying

for microcredit loans.

While both Cr-FAs and Soc-FAs increase new CO formation, among Soc-FAs this does not come

at the cost of compromising other CO quality measures. The impacts of credit bonus on rest of

the reported CO quality outcomes (Columns 2-6) are negative and large in magnitudes. Cr-FAs

decreased the share of savers in CO meetings by 12.7 percentage points, worsened attendance by

10.5 percentage points, and reduced the share of COs with multiple meetings in a month by 19.4

percentage points relative to control FAs. The estimates are statistically different from zero at the

5 percent level, and amount to 18.2, 13.5, and 46.0 percent decline in savings, attendance, and

meetings, respectively. This suggests a large negative effect of the credit bonus on the Cr-FAs’

performance in social mobilization tasks, perhaps due to a shift in their labor allocation and focus

on the incentivized microcredit-oriented outcomes. We examine these specific channels later in the

section.

In contrast, we find no change in these measures of CO quality among Soc-FAs. The impacts

of social bonus on the share of savers among CO members and their attendance in CO meetings,

which make up the remaining triggers for the social bonus, are small (-0.03 and -0.027 respectively)

13

and not statistically different from zero at conventional levels. But they are statistically different

from that of Cr-FAs (at 10 and 5 percent levels). It is worth noting that attendance levels among

control FAs are so high (close to 80 percent) that Soc-FAs would have qualified for a positive bonus

amount conditional on meeting their first trigger targets without any change in attendance. For

non-incentivized social outcomes in Columns 4-6, the performance of Soc-FAs is not significantly

different from that of control FAs.

We construct a CO quality index similar to the microcredit index, by taking an equally weighted

average of all six measures of CO quality standardized with mean zero and standard deviation one.12

The effect of social bonus on the CO quality index in Column 7 is slightly negative (-0.051 standard

deviations) but not statistically different from zero. On the other hand, the impact of credit bonus

on this index is large and negative, and statistically significantly different from zero at the 1 percent

level. Cr-FAs also performed worse on the CO quality index as compared to Soc-FAs (statistically

different at the 1 percent level).

Overall, these results suggest that the credit bonus improved the NRSP’s microcredit program,

albeit for outcomes directly incentivized, at least during the 15 month period when the bonus was

active. The credit bonus however also worsened the quality of COs thus undermining NRSP’s goal of

empowering communities through social mobilization. In comparison, the social bonus had relatively

more muted impacts on CO quality and on microcredit outcomes. It increased new CO formation

without worsening CO quality and more importantly, without adversely affecting microcredit out-

comes.

4.3 Intervening channels

To shed light on the possible mechanisms explaining the above results, we examine how the bonus

influences FA choices and shapes their work environment, both of which are integral to their overall

performance on mission objectives. These results are presented in Table 4, Panel A.

4.3.1 Effort and time allocation across tasks

Column 1 examines the impact of bonus on FA effort measured by self-reported daily overtime hours

from the follow up survey. Both credit and social bonus did not considerably increase effort. This

is not surprising since overtime work was already high (2.59 hours per day among control FAs).

Soc-FAs increased their overtime hours by 0.182, but this is not statistically significantly different

from zero. It is also not statistically different from that of Cr-FAs, who worked 0.148 hours per day

12While calculating the CO-quality index, the sign of one of the outcomes, i.e. “Dead COs,” is reversed so that forall the outcomes, positive values represent “good” outcomes.

14

more than control FAs.

In Column 2, we examine the impact of bonus on the amount of time allocated between micro-

credit (new loan and recovery) and social mobilization (attending CO meetings). Cr-FAs increased

their share of time spent on microcredit-related tasks by 12.3 percentage points, which amounts to

17.2 percent increase over the mean in the control group. This estimate is statistically significant at

the 10 percent level. It is also different from the share of time devoted by Soc-FAs to microcredit

tasks at the 5 percent level of statistical significance. Soc-FAs however do not show any significant

change in their time allocation between the two types of tasks.

Despite no increases in the overall time spent on the job, the credit bonus increased the time spent

by Cr-FAs on microcredit-related tasks. This reallocation of FA’s time towards microcredit and away

from social mobilization helps explain the differential impact of the credit bonus on microcredit and

CO quality outcomes among Cr-FAs.

4.3.2 Selection of new clients

In Column 3 we examine the impact of the bonus schemes on the perception of FA’s about who

would make a good NRSP client. The dependent variable comes from an open-ended question asked

in the follow-up survey about the qualities the FA looked for when assessing prospective clients. The

answers were post-coded in a categorical variable with values 1, -1, or 0 depending on whether the

focus was more on repayment, community empowerment, or equally on both, respectively.

While the social bonus did not change the perception towards new clients of Soc-FAs, Cr-FAs

were more likely to find themselves preferring clients with better repayment capacity rather than

clients that needed social empowerment. For Cr-FAs this effect is statistically significant at the 10

percent level relative to control FAs and significant at the 5 percent level relative to Soc-FAs.

The shift in client selection by Cr-FAs is also consistent with their improved performance on

microcredit outcomes. But more importantly, the results highlight a greater emphasis among Cr-FAs

to view CO members as potential clients for the microcredit program rather than for the broader

purpose of community empowerment, and potentially restricting NRSP’s outreach to poorer and

more needy clients.

4.3.3 Crowding out of intrinsic motivation and its impact on teamwork

Next, we examine the impact of bonus on FA’s intrinsic motivations. In the baseline and follow up

surveys, FAs were asked to list the things that they liked most about working with NRSP. Column

4 reports our measure of intrinsic motivation, which is a dummy variable that equals one if the FA

mentioned the ability to help people as what they like most from NRSP.

15

We find that in the follow-up survey, Soc-FAs are 23.4 percentage points less likely to report

being intrinsically motivated when working with NRSP compared to control FAs. This estimate is

statistically significant at the 1 percent level, indicating a substantial decline in intrinsic motivation.

Cr-FAs also show a decline in their intrinsic motivation but the estimate is not statistically different

from zero. The crowding out effect in intrinsic motivation for Cr-FAs is smaller (almost half in

magnitude) compared to Soc-FAs, but this difference is not statistically significant at conventional

levels.

While both types of monetary incentives crowd out intrinsic motivation, we do not observe a

decline in overall effort, perhaps suggesting that the extrinsic motivation provided by the bonus

scheme on average counters the decline in intrinsic motivation. For FAs assigned to teams however,

the decline in intrinsic motivation could harm productivity through the increased risk of free-riding,

especially if effort cannot be easily monitored.

Column 5 examines the impact on FA’s propensity to work in a team. We use the same variable

as in Table 1, and consider that an FA is assigned to a team (and hence is a partnered FA) if the

share of COs that are co-managed with other FAs during the pre-treatment months exceeds the

median value of (73 percent of COs).

We find that the social bonus discourages teamwork, as the share of Soc-FAs working in teams

declined by 12.5 percentage points compared to control FAs. This negative impact on teamwork is

statistically significant at the 1 percent level, and implies a roughly 20 percent decline in the share

of FAs working in teams compared to the mean in the control group (67.2 percent). This effect

is also statistically different from Cr-FAs at the 5 percent level. On other hand, credit bonus has

no impact on teamwork, with Cr-FAs slightly increasing their propensity to work in teams by 1.22

percentage points; and the estimate is not statistically different from zero at conventional levels.

In Column 6, we construct a measure of free riding in teamwork based on a dictator game played

by FAs in the follow up survey. Each FA was asked to split the winnings received from another

randomly chosen and unknown FA with this sending FA. The average winning from this game was

PKR 58.52. The dummy dependent variable Shares with a partner equals one if an FA shared some

positive amount of the total winning with his partner, and zero otherwise. The effects of credit and

social bonus on free riding are small in magnitude, and the estimates are not statistically different

from zero and from each other at conventional levels.

This is perhaps not surprising, given that the effects of free riding are likely to be concentrated

among FAs who were assigned to a team. Indeed, a decline in cooperation is more likely to occur

among partnered Soc-FAs than Cr-FAs for two main reasons. First, Soc-FAs experienced a larger

decline in intrinsic motivation than Cr-FAs; and in so far as intrinsic motivation limits free riding

16

in teams, the partnered Soc-FAs are relatively more likely to exhibit greater propensity to free ride.

Second, Soc-FAs are incentivized on more complex tasks for which inferring the partner’s effort is

more difficult. As a result, even in the presence of peer pressure, noisier indicators of effort can

create incentives to free ride.

In Panel B, Column 6 we estimate the impact of credit and social bonus on free riding for

FAs assigned to teams versus those working independently in the baseline. Among partnered FAs,

the social bonus more than doubled the share of FAs who did not spilt the winnings with their

partners in a dictator game, suggesting a considerable increase in free riding among partnered Soc-

FAs (statistically significant at the 10 percent level). For non-partnered FAs however the impact of

social bonus on cooperative behavior was positive and also statistically significant at the 10 percent

level. The within-treatment difference in free riding between partnered and non-partnered FAs with

the social bonus is statistically significant at the 5 percent level. Additionally, we do not observe

changes in free riding behavior among partnered and non-partnered Cr-FAs.

In sum, the results on teamwork suggest that the crowding out of intrinsic motivation due to

monetary bonuses, negatively affect the ability of Soc-FAs to perform in teams by worsening the

free-riding problem. Next, we explore whether this adverse effect on teamwork also translates into

poorer performance.

4.4 Performance differences by baseline partnership

In Panel B of Tables 2 and 3, we estimate the impacts of credit and social bonus on the microcredit

and CO-quality outcomes separately for partnered and non-partnered FAs. We highlight the hetero-

geneity in performance on microcredit outcomes in Table 2. For non-partnered Cr-FAs, the effects

on the two trigger outcomes are large and positive, and the effect on the repayment at 20th of the

month is also statistically significant at the 1 percent level. On the microcredit index (in Column

6), non-partnered Cr-FAs performed 0.277 standard deviations better than non-partnered control

FAs (statistically significant at the 10 percent level). Their performance on this index is however

not statistically different from that of non-partnered Soc-FAs at conventional levels.

Non-partnered Soc-FAs also performed 0.297 standard deviations better on the microcredit index

compared to control FAs. This estimate is statistically significant at the 10 percent level. In fact,

they performed better than their control counterparts on all the reported microcredit outcomes,

unlike non-partnered Cr-FAs for whom the improvements were mainly focused on the two trigger

outcomes. As a result, non-partnered Soc-FAs even outperformed Cr-FAs on various disbursement

measures (active loans, new loans, and disbursement amount) and the difference is statistically

significant (at the 10 percent level) for the number of new loans.

17

Next, we examine the effects on partnered FAs. In line with our priors, partnered Soc-FAs

performed extremely poorly on all microcredit outcomes. Partnered Soc-FAs decreased repayment

at 20th of the month by 7.3 percentage points, decreased repayment at the end of the month by

0.25 percentage points; and they performed 0.269 standard deviations lower in the microcredit index

compared to partnered control FAs. These effects are statistically significant at the 1 and 10 percent

levels. On the microcredit index, the coefficient on the interaction term TSxP is negative (-0.566

standard deviations) and statistically significant at the 1 percent level. In contrast, there are no

sizable within treatment differences by partnership among Cr-FAs. For all the microcredit outcomes,

the coefficient on the interaction term TCxP is not statistically significantly different from zero at

conventional levels. On the microcredit index, the TCxP interaction term is small in magnitude

(-0.081 standard deviations), and not statistically significant.

This pattern of heterogeneity is also observed with CO quality in Table 3. We find that partnered

Soc-FAs perform considerably worse than non-partnered Soc-FAs in all (expect one) CO quality

outcomes. On the CO quality index, the coefficient on TSxP is -0.304 standard deviations, and this

estimate is statistically significant at the 10 percent level. As a result, among partnered FAs, social

bonus lowered performance on the CO quality index by 0.182 standard deviations, while it improved

performance of non-partnered FAs by 0.122 standard deviations. The former estimate is statistically

significant at the 10 percent level. In contrast, there are no differences in the credit bonus between

partnered and non-partnered FAs. On the CO quality index, the coefficient on TCxP is close to zero

(0.015 standard deviations) and it is not statistically significant at conventional levels.

In sum, partnership among FAs assigned to the credit bonus is mostly benign, but it has a large

and deleterious effect on Soc-FAs. The muted effect of social bonus on microcredit and CO quality

outcomes (in Panel A of Tables 2 and 3) is partly due to this negative effect among partnered

Soc-FAs. Among non-partnered FAs, the social bonus improved both microcredit and CO quality

outcomes. The results suggest that among non-partnered FAs, the social bonus was as effective as

the credit one on improving microcredit outcomes. Therefore, incentivizing social mobilization, at

least when teamwork is not involved, is likely to complement NRSP’s microcredit program, while

the converse is not true. On the other hand, social bonus under partnership is as bad as the negative

spillovers created by the credit bonus on CO-quality.

4.5 Effects after bonus ended

While the bonus incentives were terminated after 15 months in June 2006, introducing monetary

incentives for a fixed period of time might affect outcomes in the long run. This could occur if

such temporary incentives change the perceived reference point in FA’s employment contract, or if

18

FAs no longer view the incentivized tasks as being pro-social and therefore cannot derive intrinsic

motivation from performing the tasks (Gneezy and Rustichini, 2000). In addition, the incentives

can directly affect outcomes that generate positive externalities in the future.

To evaluate the performance of Cr-FAs and Soc-FAs in the post-bonus period, we construct a

new FA-CO panel from July 2006 to November 2006. First, we take the FAs who were still employed

with NRSP in June 2006. Then, we construct their post-bonus CO-portfolio based on whether they

were the last FAs to manage the CO in the original FA-CO panel (during the bonus period). We take

the MIS data from July 2006 to November 2006, and merge the monthly CO microcredit outcomes

with this new FA-CO panel. Lastly, we aggregate the microcredit outcomes for each FA by month,

and then calculate the post-bonus performance of an FA by taking the average over the 5 post-bonus

months. For each FA, we use the same pre-treatment performance as in the earlier analysis (the

average across 9 pre-bonus months).

Out of the original 162 FAs, we have information on their post-bonus CO-portfolio for 136 FAs.

Table 1 (bottom panel) provides the selection rate for control FAs, Cr-FAs, and Soc-FAs. The

selection rate is almost identical and not statistically different across the three groups. In addition,

Appendix Table D.3 presents means of baseline variables for control FAs, Cr-FAs, and Soc-FAs in

this restricted post-bonus sample, and the P-values from the F-tests of the differences in the means

between the three groups. Three out of 69 differences in means are statistically significant at the 10

percent level. The F-tests in Columns 4-6 do not reject the joint equality of means between control

and Cr-FAs, control and Soc-FAs, and Cr-FAs and Soc-FAs respectively.

Within five months following the completion of the bonus experiment, all the improvements we

observed among the Cr-FAs disappear, so that they no longer look any better than the control

FAs on the microcredit outcomes. Based on the results from Table 5, Panel A, Column 6, Cr-FAs

were performing only 0.012 standard deviations better on the microcredit index than control FAs

(compared to 0.238 standard deviations when bonus was active), and this estimate is not statistically

different from zero. Their performance on this index is also not statistically different from that of

Soc-FAs. This suggests that the positive impact of credit bonus on NRSP’s microcredit program

was temporary and lasted only till the bonus was active.

The social bonus, in contrast, continued to affect the performance of Soc-FAs beyond the timeline

of the bonus experiment. Among non-partnered Soc-FAs, we find sizable improvements on their

active loans, new loans, and disbursement during the post-bonus months, while partnered Soc-FAs

continue to underperform in terms of expanding their microcredit loan portfolio (active loans, new

loans, and new disbursement). These effects are statistically significant at the 5 and 10 percent

levels. The results for partnered and non-partnered Soc-FAs are in line with how each of them

19

performed on their CO-quality outcomes during the bonus period. It provides strong suggestive

evidence that forming better quality COs through the process of social mobilization also make for

stronger clients for microcredit program.

5 Conclusion

In NRSP, a nonprofit development organization in Pakistan, we implemented two types of monetary

performance-based incentives for its workers to test the potential trade-offs between its microcredit

program and its overarching goal of social and economic empowerment. The credit bonus, which

incentivized more short-term, bottom-line goals directly related to the microcredit operation, led to

immediate improvement in loan disbursements and repayments; but at a serious cost of worsening

CO-quality, therefore undermining its main mission of empowerment. In contrast, the social bonus,

which incentivized social mobilization goals that are likely to be more aligned with the overall

mission, was as effective as the credit bonus in improving the microcredit outcomes among non-

partnered FAs, and it improved their CO formation without compromising CO-quality. Any positive

impact of credit bonus disappeared after the bonus ended, while investing in social mobilization had

long-term pay-off for its microcredit operation. These results reveal the importance of choosing the

appropriate incentive structures (in relation to the choice of goals to incentivize on), which influences

the overall effectiveness of such performance-based incentives in achieving organization’s mission.

From a policy standpoint, there is a substantial emphasis on the commercialization of develop-

ment organizations that also provide microcredit to ensure their long-term sustainability. Our results

highlight the main challenge that these organizations face in this context, where an overemphasis on

bottom-line goals might not only undermine its overarching development goals, but also affect its

long-term sustainability as a credit institution by restricting its future pool of creditworthy clients.

Our results also reveal that importance of intrinsic motivation in ensuring the quality of the

services, which more often than not extend beyond microcredit operation, that are offered by these

organizations. The interconnectedness in terms of how these services are delivered, might imply that

workers might not be able to separate out the motivation behind each individual tasks. In our study

both types of incentives led to a decline in intrinsic motivation across the board.

There are important aspects of mission-oriented organization that are not explored in this paper,

in particular the role of how incentives match with worker’s mission-preferences, as well as the effect

of incentives on the relationship between FAs and their supervisors. A detailed examination of these

two factors is an important area for future research, which we intend to explore.

20

References

Baker, George, “Incentive Contracts and Performance Measurement,” Journal of Political Econ-omy, 1992, 100 (3), 598-614.

Bandiera, Oriana and Iwan Barankay and Imran Rasul, “Team Incentives: Evidence From AFirm Level Experiment,” Journal of the European Economic Association, 2013, 11(5), 1079-1114.

Bandiera, Oriana and Iwan Barankay and Imran Rasul, “Incentives for Managers and In-equality Among Workers: Evidence From a Firm-Level Experiment,” The Quarterly Journal ofEconomics, 2007, 122(2), 729-773.

Basinga, Paulin, Paul J. Gertler, Agnes Binagwaho, Agnes L. B. Soucat, JenniferSturdy, and Christel M. J. Vermeersch, “Effect on Maternal and Child Health Services inRwanda of Payment to Primary Health-Care Providers for Performance: An Impact Evaluation,”Lancet, 2011, 377 (9775), 1421-1428.

Besley, Timothy and Maitreesh Ghatak, “Competition and Incentives with Motivated Agent,”The American Economic Review, 2005, 95 (3), 616-636.

Bo, Ernesto Dal, Frederico Finan, and Martin A. Rossi, “Strengthening State Capabili-ties: The Role of Financial Incentives in the Call to Public Service,” The Quarterly Journal ofEconomics, 2013, 128 (3), 1169-1218.

Bowles, Samuel and Sandra Polanina-Reyes, “Economic Incentives and Social Preferences:Substitutes or Complements?,” Journal of Economic Literature, 2012, 50(2), 368-425.

Duflo, Esther, Rema Hanna and Stephen P. Ryan, “Incentives Work: Getting Teachers toCome to School,” The American Economic Review, 2012, 102(4), 1241-1278.

Glewwe, Paul, Nauman Ilias, and Michael Kremer, “Teacher Incentives,” American Eco-nomic Journal: Applied Economics, 2010, 2, 205-227.

Gneezy, Uri and Aldo Rustichini, “A Fine Is a Price,” The Journal of Legal Studies, 2000,29(1), 1-17.

Goodman, Sarena F. and Lesley J. Turner, “The Design of Teacher Incentive Pay and Educa-tional Outcomes: Evidence from the New York City Bonus Program,” Journal of Labor Economics,2013, 31(2).

Government of Pakistan, “Federal Budget in Brief,” Finance Division, 2014.

Holmstrom, Bengt and Paul Milgrom, “Multitask Principal-Agent Analyses: Incentive Con-tracts, Asset Ownership, and Job Design,” Journal of Law, Economics, Organization, 1991, 7,24-52.

Imberman, Scott A. and Michael F. Lovenheim, “Incentive Strength and Teacher Productiv-ity: Evidence from a Group-Based Teacher Incentive Pay System,” The Review of Economics andStatistics, 2015, 97(2), 364-386.

Kling, Jeffrey, Jeffrey Liebman, and Lawrence Katz, “Experimental Analysis of Neighbor-hood Effects,” Econometrica, 2007, 75 (1), 83-119.

Lavy, Victor, “Evaluating the Effect of Teachers’ Group Performance Incentives on Pupil Achieve-ment,” Journal of political Economy, 2002, 110, 1286-1317.

21

Lavy, Victor, “Performance Pay and Teachers’ Effort, Productivity, and Grading Ethics,” Ameri-can Economic Review, 2009, 99, 1979-2011.

Lazear, Edward P., “Performance Pay and productivity,” The American Economic Review, 2000,90(5), 1346-1361.

Masuri, Ghazala and Vijendra Rao, Localizing Development: Does Participation Work?, TheWorld Bank: Washington, D.C., 2000.

Montagu, Dominic and Gavin Yamey, “Pay-for-performance and the Millennium DevelopmentGoals,” Lancet, 2011, 377 (9775), 23-29.

Muralidharan, Karthik and Venkatesh Sundararaman, “Teacher Performance Pay: Experi-mental Evidence from India,” Journal of Political Economy, 2011, 119(1), 39-77.

Murphy, Kevin J., “Executive Compensation,” Handbook of Labor Economics, 1999, 3 (B), 2485-2563.

National Rural Support Program, “20th Annual Progress Report: Managing Scale,” 2014.

Olken, Benjamin A., Junko Onishi, and Susan Wong, “Should Aid Reward Performance?Evidence from a Field Experiment on Health and Education in Indonesia,” American EconomicJournal: Applied Economics, 2014, 6(4), 1-34.

Osterloh, Margit and Bruno S. Frey, “Motivation, Knowledge Transfer, and OrganizationalForms,” Organization Science, 2000, 11 (5), 538-550.

Paarsch, and Bruce Shearer, “Piece Rates, Fixed Wages, and Incentive Effects: StatisticalEvidence from Payroll Records,” International Economic Review, 2000, 41(1), 59-92.

Prendergast, Canice, “The Provision of Incentives in Firms,” Journal of Economic Literature,1999, 37(1), 7-63.

Shearer, Bruce, “Fixed Rates, Fixed Wages and Incentives: Evidence from a Field Experiment,”Review of Economic Studies, 2004, 71, 513-534.

World Bank, “Pakistan Country Snapshot,” October 2014.

22

Table 1: Summary statistics and balance tests

No Credit Socialbonus bonus bonus P-values(C) (TC) (TS) C=TC C=TS TC=TS(1) (2) (3) (4) (5) (6)

Demographic characteristicsAge 27.39 27.53 28.03 0.928 0.646 0.765Female 0.266 0.250 0.129 0.879 0.105 0.139Married 0.375 0.361 0.435 0.910 0.524 0.540Household head 0.125 0.167 0.177 0.558 0.498 0.891Completed high school 0.562 0.500 0.565 0.552 0.979 0.543Household consumption (PKR) 6531 5875 6874 0.396 0.539 0.151Owns land 0.422 0.444 0.548 0.870 0.301 0.456Housing quality index 0.167 -0.094 0.265 0.131 0.595 0.109

Employment characteristicsWork from a village branch 0.781 0.889 0.903 0.367 0.286 0.863Employed with NRSP (months) 26.92 25.97 26.40 0.673 0.859 0.895Number of COs managed 12.29 16.59 14.62 0.274 0.336 0.621Share of COs co-managed 0.565 0.494 0.597 0.423 0.681 0.253

Partnered FAa 0.500 0.389 0.565 0.408 0.657 0.248Overtime work (hrs/day) 1.940 1.748 2.223 0.590 0.369 0.172Time spent on microcredit tasks (%) 0.728 0.753 0.811 0.733 0.113 0.413Prefers credit bonus 0.594 0.528 0.565 0.612 0.819 0.779Social mobilization helps microcredit 0.906 0.944 0.952 0.392 0.309 0.867Did volunteer work before NRSP 0.250 0.194 0.290 0.569 0.671 0.378Best about NRSP: ability to help 0.484 0.528 0.565 0.611 0.404 0.732

Monthly performanceNumber of active loans 75.19 99.87 103.3 0.338 0.239 0.897New disbursement (PKR) 61894 100675 92354 0.108 0.107 0.753Repayment on dues at 20th of month 0.749 0.724 0.703 0.655 0.431 0.774Repayment on dues at end of month 0.984 0.968 0.994 0.566 0.234 0.347Number of field units (FUs) 11 9 15 - - -Number of field assistants (FAs) 64 36 62 - - -Number of credit organizations (COs) 1411 1217 1776 - - -FA attrition in followup survey 0.156 0.194 0.210 0.721 0.586 0.876FA selection into verified MPRsτ 0.719 0.944 0.823 0.186 0.564 0.233FA selection into post-bonus FA-COζ 0.797 0.861 0.806 0.463 0.901 0.531F-test statistics - - - 0.959 1.033 1.207P-value - - - 0.523 0.433 0.268

Notes: The standard errors (in F-tests in Columns 4-6) are adjusted for within-field unit correlation between FAs.aPartnered FA is defined as a dummy variable which equals one if an FA is co-managing more than 73 percent of her/hisCO portfolio (median value of co-sharing) with other FAs during the 9 months prior to the bonus announcement.τVerified MPRs sample include FAs whose CO meetings were visited by the CrO at least once during the bonus period.ζPost-bonus FA-CO sample includes FAs who were still working with NRSP and were managing at least one CO at the endof June 2006.

23

Table 2: Impact of bonus on microcredit

Bonus triggersNumber Repayment New New Repayment Microcreditof active on dues at loans disburse- on dues at index

loans 20th of month ment end of month(1) (2) (3) (4) (5) (6)

Panel A: average treatment effect

Credit bonus (TC) 24.1** 0.084** 0.673 -284.0 0.008 0.238*(11.500) (0.0399) (1.5440) (23534) (0.0121) (0.1380)

Social bonus (TS) 9.406 -0.020 0.616 -444.8 -0.0034 0.002(14.510) (0.0367) (1.8980) (25580) (0.0105) (0.1490)

P-value of F-test:TC = TS 0.285 0.020 0.966 0.993 0.283 0.072

R-squared 0.544 0.566 0.383 0.399 0.364 0.449

Panel B: by level of co-management

Credit bonus (TC) 14.88 0.117*** -0.429 -9801 0.023 0.277*(17.030) (0.0385) (1.7360) (27854) (0.0161) (0.1400)

Social bonus (TS) 27.26 0.037 3.389 29389 0.020 0.297*(22.190) (0.0367) (2.3540) (31810) (0.0183) (0.1720)

Partnered FA (P) -18.92 0.031* -3.214** -40066** 0.034* -0.049(13.720) (0.0164) (1.3190) (18230) (0.0176) (0.1020)

P x TC 22.76 -0.069 2.729 22211 -0.027 -0.081(23.620) (0.0471) (1.7670) (24895) (0.0226) (0.1460)

P x TS -33.18 -0.110*** -5.150* -55657 -0.045* -0.566***(26.750) (0.0386) (2.5540) (33567) (0.0223) (0.1870)

P-value of F-test:TC + P x TC 0.0184 0.396 0.198 0.591 0.823 0.267TS + P x TS 0.676 0.0756 0.300 0.254 0.0568 0.0490TC = TS 0.640 0.0476 0.090 0.158 0.816 0.909TC+PxTC=TS+PxTS 0.000 0.077 0.004 0.038 0.212 0.008

R-squared 0.572 0.589 0.463 0.460 0.405 0.508

Observations 162 162 162 162 162 162Mean dep. var., control 118.8 0.716 11.00 138347 0.964 -0.163

Notes: All specifications control for region dummies and a pre-treatment value of the dependent variable. Partnered FA is adummy variable which equals one if an FA co-manages more than 73 percent of her/his pre-treatment CO portfolio (medianvalue of co-sharing) with other FAs. New COs is the monthly average number of active loans (new and on-going) managedby the FA. Repayment on dues at 20th of the month is the monthly average share of installment dues paid in full by the 20th.New loans is the monthly average number of new loans issued by the FA. New disbursement is the monthly average amountof new loans issued by the FA in Rupees. Repayment on dues at end of month is the monthly average share of installmentdues that were paid in full by end of the month. Microcredit index is calculated by taking an equally weighted mean acrossthe standard distributions of the five microcredit outcomes in Columns 1-5. Higher value on the microcredit index impliesbetter performance on microcredit. The standard errors are reported in parentheses and are adjusted for within-field unitcorrelation between FAs; *p<0.1, **p<0.05, ***p<0.01.

24

Table 3: Impact of bonus on CO quality

Bonus triggersNew Savers Attend- Dead Multiple Loan CO-qualityCOs per ance COs meetings rejection index

member rate(1) (2) (3) (4) (5) (6) (7)


Credit bonus (TC) 0.284* -0.127** -0.105** 1.020 -0.194** -0.0793 -0.384***(0.1470) (0.0529) (0.0387) (0.6820) (0.0946) (0.0568) (0.1180)

Social bonus (TS) 0.225** -0.030 -0.027 0.056 -0.040 -0.047 -0.051(0.0936) (0.0376) (0.0372) (0.3220) (0.0909) (0.0460) (0.1010)

P-value of F-test:TC = TS 0.696 0.079 0.013 0.117 0.139 0.415 0.004

R-squared 0.230 0.271 0.286 0.430 0.293 0.096 0.442


Credit bonus (TC) 0.181 -0.121** -0.084** 1.494 -0.168 -0.054 -0.382***(0.1690) (0.0498) (0.0343) (1.0380) (0.1280) (0.0613) (0.1240)

Social bonus (TS) 0.430** -0.059 0.012 -0.326 0.013 -0.024 0.122(0.1590) (0.0556) (0.0344) (0.6630) (0.1300) (0.0678) (0.1360)

Partnered FA (P) -0.173* -0.080 -0.039 -0.824 0.023 0.004 -0.113(0.0957) (0.0733) (0.0603) (0.6550) (0.0646 (0.0381) (0.1420)

P x TC 0.264 -0.028 -0.047 -1.30 -0.050 -0.057 0.015(0.1890) (0.0997) (0.0725) (1.0360) (0.1550) (0.0545) (0.2470)

P x TS -0.356** 0.051 -0.068 0.662 -0.096 -0.041 -0.304*(0.1660) (0.0896) (0.0663) (0.7810) (0.1400) (0.0572) (0.1760)

P-value of F-test:TC + P x TC 0.008 0.120 0.041 0.627 0.069 0.104 0.079TS + P x TS 0.236 0.898 0.278 0.275 0.424 0.145 0.098TC = TS 0.163 0.333 0.006 0.052 0.216 0.633 0.000TC+PxTC=TS+PxTS 0.032 0.099 0.089 0.607 0.300 0.194 0.322

R-squared 0.338 0.305 0.360 0.483 0.297 0.107 0.496

Observations 131 131 131 131 131 131 131Mean dep. var., control 0.384 0.699 0.777 -2.230 0.422 0.138 0.118

Notes: All specifications control for region dummies and a pre-treatment value of the dependent variable. Partnered FA is adummy variable which equals one if an FA co-manages more than 73 percent of her/his pre-treatment CO portfolio (medianvalue of co-sharing) with other FAs. New COs is the monthly average number of new COs formed by the FA. Savers permember is the monthly average share of CO members who saved during CO meetings conducted by the FA. Attendance isthe monthly average share of CO members present at the CO meetings conducted by the FA. Dead COs is the monthlyaverage number of COs managed by the FA without any active borrowers for the entire bonus period. Multiple meetings isthe monthly average share of COs managed by the FA that had more than one monthly meetings. Loan rejection rate is themonthly average share of social appraisals rejected by the FA. CO-quality index is calculated by taking an equally weightedmean across the standard distributions of the six Co-quality outcomes in Columns 1-6. Higher value on the CO-quality indeximplies better performance on social mobilization. The standard errors are reported in parentheses and are adjusted forwithin-field unit correlation between FAs; *p<0.1, **p<0.05, ***p<0.01.

25

Table 4: Impact of bonus on FA choice and work environment

Overtime % time Credit> Best about Works Shareswork spent on social for NRSP: ability with a with a

(hrs/day) microcredit new clients to help partner partner(1) (2) (3) (4) (5) (6)


Credit bonus (TC) 0.148 0.123* 0.157* -0.141 0.012 -0.011(0.3030) (0.0709) (0.0861) (0.1020) (0.0759) (0.0410)

Social bonus (TS) 0.182 0.015 -0.039 -0.234*** -0.125*** -0.012(0.1930) (0.0620) (0.0898) (0.0849) (0.0447) (0.0327)


R-squared 0.333 0.334 0.025 0.137 0.570 0.058


Credit bonus (TC) -0.099 0.152 0.113 -0.185 -0.09 0.030(0.2560) (0.0941) (0.1330) (0.1650) (0.1520) (0.0575)

Social bonus (TS) 0.331 -0.062 -0.085 -0.177 -0.164 0.078*(0.2200) (0.1010) (0.1080) (0.1470) (0.1360) (0.0440)

Partnered FA (P) -0.338 -0.025 -0.016 0.009 0.514*** 0.081*(0.3250) (0.0970) (0.0784) (0.1300) (0.1830) (0.0460)

P x TC 0.555 -0.084 0.096 0.116 0.042 -0.071(0.5300) (0.1160) (0.2250) (0.2510) (0.2160) (0.1220)

P x TS -0.271 0.149 0.088 -0.109 0.074 -0.170**(0.3830) (0.1330) (0.1480) (0.1690) (0.2030) (0.0657)


R-squared 0.355 0.353 0.027 0.145 0.353 0.079

Observations 132 132 132 132 162 132Mean dep. var., control 2.593 0.716 0.148 0.500 0.716 0.944

Notes: All specifications control for region dummies. All specifications control for a pre-treatment value of the dependentvariable (except Columns 4 and 7). Partnered FA is a dummy variable which equals one if an FA co-manages more than 73percent of her/his pre-treatment CO portfolio (median value of co-sharing) with other FAs. Overtime is the average numberof overtime hours per day self-reported by the FA. % time spent on credit tasks is the ratio of time spent each month byan FA on loan recovery and social appraisals over the total time spent on the two credit tasks plus attending CO meetings.Credit>social for new clients is a category variable which takes the values of 1, -1, and 0 depending on if an FA reportsthat qualities related to repayment is more important when looking for new clients to form COs, social mobilization qualityis important, or both is important, respectively. Best about NRSP: ability to help is a dummy variable which equals oneif an FA ranks the ability to help people as the best thing about working with NRSP. Shares with a partner is a dummyvariable which equals one if an FA decides to split the winnings with her/his colleague from the dictator game. Works witha partner is a dummy variable which equals one if an FA co-manages more than 73 percent of her/his post-treatment COportfolio with other FAs.The standard errors are reported in parentheses and are adjusted for within-field unit correlationbetween FAs; *p<0.1, **p<0.05, ***p<0.01.

26

Table 5: Impact of bonus on microcredit (after bonus ended)

Bonus triggersNumber Repayment New New Repayment Microcreditof active on dues at loans disburse- on dues at index

loans 20th of month ment end of month(1) (2) (3) (4) (5) (6)


Credit bonus (TC) 23.56 -0.016 2.250 21132 -0.023 0.012(21.080) (0.0430) (1.8600) (25371) (0.0234) (0.1270)

Social bonus (TS) 9.53 -0.072 2.956 34104 -0.036 -0.072(23.750) (0.0442) (2.5860) (34620) (0.0302) (0.1620)


R-squared 0.315 0.351 0.332 0.326 0.361 0.451


Credit bonus (TC) 27.11 -0.053 3.140 38405 -0.044 0.020(33.140) (0.0499) (2.1800) (28325) (0.0274) (0.1440)

Social bonus (TS) 50.52* -0.071 7.762** 95,187** -0.049 0.163(29.570) (0.0568) (3.2570) (42108) (0.0427) (0.2070)

Partnered FA (P) -4.32 -0.023 0.069 1,478 -0.008 -0.035(14.980) (0.0387) (1.2140) (15343) (0.0202) (0.1020)

P x TC -5.08 0.089 -1.878 -38511 0.049 -0.002(56.530) (0.0600) (2.8490) (35856) (0.0297) (0.2440)

P x TS -85.48** -0.005 -10.11** -128677** 0.027 -0.494**(32.120) (0.0596) (3.7070) (47758) (0.0347) (0.2290)


R-squared 0.356 0.360 0.394 0.386 0.374 0.487

Observations 136 136 136 136 136 136Mean dep. var., control 129.3 0.754 5.721 83169 0.960 -0.169

Notes: All specifications control for region dummies and a pre-treatment value of the dependent variable. Partnered FA is adummy variable which equals one if an FA co-manages more than 73 percent of her/his pre-treatment CO portfolio (medianvalue of co-sharing) with other FAs. New COs is the monthly average number of active loans (new and on-going) managedby the FA. Repayment on dues at 20th of the month is the monthly average share of installment dues paid in full by the 20th.New loans is the monthly average number of new loans issued by the FA. New disbursement is the monthly average amountof new loans issued by the FA in Rupees. Repayment on dues at end of month is the monthly average share of installmentdues that were paid in full by end of the month. Microcredit index is an equally weighted average across the standarddistributions of the five microcredit outcomes in Columns 1-5. Positive values of microcredit index imply better performanceon microcredit. The standard errors are reported in parentheses and are adjusted for within-field unit correlation betweenFAs; *p<0.1, **p<0.05, ***p<0.01.

27

Figure 1: Monthly payments of credit and social bonus

Credit bonus

Social bonus

0.2

.4.6

.81

%

May05 Jul05 Sep05 Nov05 Jan06 Mar06 May06

(a) Qualified for a bonus

0.2

.4.6

.81

%


(b) Target A

0.2

.4.6

.81

%


(c) Target B

0100

200

300

400

500

600

PKR


(d) Bonus amount

28

A Appendix: Study Design and Bonus Incentives

Table A.1: Description of the credit and social bonus incentives

Panel A: Credit bonus

The first trigger is based on disbursement, measured by the number of active loans managed by the FA in anymonth. The second trigger is based on whether the repayment on the installment was made in full by the 20 th

of the month due. The disbursement trigger can be satisfied at two target levels: High (A) or Low (B). If theFA meets at least target B for disbursement, he qualifies for a bonus based on his recovery rate at the 20 th inthat month.

If FA qualifies on target A, the size of the bonus is:20% of base monthly salary if repayment is 100%16% of base monthly salary if repayment is 99%12% of base monthly salary if repayment is 98%8% of base monthly salary if repayment is 97%4% of base monthly salary if repayment is 96%0 bonus if repayment is 95% or below

If FA qualifies on target B, the size of the bonus is:15% of base monthly salary if repayment is 100%12% of base monthly salary if repayment is 99%9% of base monthly salary if repayment is 98%6% of base monthly salary if repayment is 97%3% of base monthly salary if repayment is 96%0 bonus if repayment is 95% or below

The bonus cannot ever be negative.

Panel B: Social bonus

The first trigger is based on two outcomes: the number of new COs formed and the number of savers at COmeetings. High (A) and Low (B) target levels are set for both outcomes, and an FA needs to reach at leasttarget B for both outcomes to satisfy the first trigger. The second trigger is based on the attendance of COmembers at CO meetings. If an FA meets at least target B, he qualifies for a bonus based on member attend-ance at CO meetings.

If the FA qualifies on target A, the size of the bonus is:20% of base salary if average attendance is 85% or more (more than 60% in harvest months)16% of base salary if average attendance is 80% to 84%(between 56% and 60% in harvest months)12% of base salary if average attendance is 75% to 79% (between 50% and 55% in harvest months)8% of base salary if average attendance is 70% to 74% (between 46% and 50% in harvest months)4% of base salary if average attendance is 65% to 69% (between 40% and 45% in harvest months)0 bonus if attendance is below 65% (0 bonus if attendance is below 40%)

If the FA qualifies on target B, the size of the bonus is determined as follows:15% of base salary if average attendance is 85% or more (more than 60% in harvest months)12% of base salary if average attendance is 80% to 84% (between 56% and 60% in harvest months)9% of base salary if average attendance is 75% to 79% (between 50% and 55% in harvest months)6% of base salary if average attendance is 70% to 74% (between 46% and 50% in harvest months)3% of base salary if average attendance is 65% to 69% (between 40% and 45% in harvest months)0 bonus if attendance is below 65% (0 bonus if attendance is below 40%)

The bonus cannot ever be negative.

Notes:The bonus incentives were announced to the FAs in the treatment FUs in Mach 2005. The monthlybonuses were paid for 15 months during the study period, and terminated in June 2006. The average basemonthly salary for an FA was PKR 3,000 (USD 50.54).

29

Table A.2: List of FUs and bonus assignments

Region District Field Unit

Panel A: No bonus(control group) Hyderabad Badin Matli

Hyderabad Badin TalharHyderabad Hyderabad HalaHyderabad Mir Pur Khas DigriHyderabad Mir Pur Khas Ghulam MuhammadHyderabad Thatta Mirpur SakroMalakand Malakand DargaiMianwali Bhakkar BhakkarMianwali Mianwali Mianwali (Swans)Rawalpindi Attock HasanabdalRawalpindi Gujar Khan Gujar Khan

Panel B: Credit bonusHyderabad Badin Badin II (Golarchi)Hyderabad Hyderabad MatiariHyderabad Mir Pur Khas HyderabadMalakand Malakand ThanaMalakand Mardan KatlangMianwali Khusab QuaidabadRawalpindi Attock AttockRawalpindi Jand JandRawalpindi Jand Pindi Gheb

Panel C: Social bonusHyderabad Badin Tando BagoHyderabad Hyderabad Tando Allah YarHyderabad Hyderabad Tando M. KhanHyderabad Thatta Mirpur BathoroHyderabad Thatta SajawalMalakand Malakand Hero ShahMalakand Malakand KabalMalakand Mardan HatianMalakand Mardan Takhat BhaiMianwali Bhakkar Dulle WalaMianwali Bhakkar MankeraMianwali Khusab JauharabadRawalpindi Attock Fateh JangRawalpindi Gujar Khan DoltalaRawalpindi Pind Dadan Pind Dadan Khan

Notes: The study was conducted in 35 Field Units (FUs) of NRSP located in 15 districts and four regionsof Pakistan, where NRSP was active in March 2005. The two treatment and control assignments wererandomly allocated across these 35 FUs. All FAs working in an FU received the same type of bonus (orcontrol group).

30

Figure A.1: Timeline of the study

June 06

Follow up Survey

Jan 05-‐Feb 05 March 05

Bonus Announced

Monthly Bonus Payments

Baseline Survey

FA-‐CO Panel and MIS data (25 months)

June 04

MPRs data (15 months)

April 05

31

B Appendix: Selection into CrO Visits

The data on CO-quality outcomes are based on the Monthly Progress Report (MPRs) filed by each

FA for all COs visited that month. The MPRs data is available for 15 months when the bonus

was implemented. This data was verified by a Credit Officer (CrO) through visits to a subset of

scheduled CO meetings. We use the verified MPRs data for the analysis.

According to the FA-CO panel, 4,380 unique COs were managed by FAs in our study sample

during the bonus period. Out of them, 1,807 COs (41.26%) were visited by a CrO at least once in

15 months, and 6 months out of 15 on average. We take all the COs that show up in the FA-CO

panel for each month (during the bonus period), and estimate the rate of CrO visits across the two

treatment and the control groups using the following specification:

CrOc,t = α + βr + ωt + γ TCc + λ TSc + ε

where CrOc,t is an indicator variable that takes the value of one if a CO c was visited by a CrO in

month t. βr and ωt are region and month dummies, respectively. The coefficients γ and λ estimate

the propensity of CrO visits to CO meetings that are managed by credit and social bonus FAs,

respectively (compared to CO meetings managed by control FAs).

Appendix Table B.1, Column 1 presents the estimated results, with standard errors clustered at

the FU level. Among control FAs, 10.5 percent of CO meetings was visited by a CrO. The estimated

coefficients γ and λ are both close to zero (0.039 and -0.000). The estimates are also not statistically

different from zero and from each other at conventional levels.

While we do not find evidence of a differential rate of CrO visits, we also test for any potential

selection on CO characteristics. For this purpose, we calculate CO’s disbursement and repayment

outcomes for each month (during the bonus period) using information from the MIS data, and then

estimate the following specification:

Yc,t = α + βr + ωt + ζ CrOc,t + γ TCc + λ TSc + φ CrOc,t ∗ TCc + ψ CrOc,t ∗ TSc + ε

where Yc,t is the characteristics of a CO c in month t. The coefficients φ and ψ on the interaction

terms CrOc,t ∗ TCc and CrOc,t ∗ TSc represent the difference in CO characteristics for those that

were visited by a CrO compared with those that were not, among COs managed by credit and social

bonus FAs, respectively (relative to the same difference among COs managed by control FAs).

Appendix Table B.1, Columns 2-6 report the results on five different CO-characteristics: the

number of active loans, number of new loans, disbursement, and recovery rates at the 20th and

32

at the end of the month. The estimated coefficient φ is close to zero (if anything, negative) and

not statistically significantly different from zero at conventional levels for all five outcomes. The

estimated coefficient ψ is also negative in sign for these five outcomes, although not statistically

significantly different from zero at conventional levels (except for number of active loans, which

is statistically significant at the 10 percent level). The two coefficients are also not statistically

significantly different from each other at the conventional levels for all five outcomes.

Column 7 presents the results on the CO-characteristics index, which is calculated by taking

an equally weighted average across the standard distributions of the five measures. The results in

Column 7 also suggest no differential selection into CrO visits to CO meetings that are managed by

credit and social bonus FAs (compared control FAs). Both coefficients are not statistically signifi-

cantly different from zero and from each other at the conventional levels. The negative signs on both

coefficients φ and ψ suggest a plausibly negative selection (if anything), which would underestimate

our main results on social outcomes (in Table 3).

33

Table B.1: Selection on frequency and the quality of CO meetings visited by a CrO

CrO Number New New Repayment Repayment CO-charac-visit of active loans disbursement on dues at on dues at teristics

loans 20th of month end of month index(1) (2) (3) (4) (5) (6) (7)

Credit bonus (TC) 0.039 0.335 0.005 -400.5 0.041 0.013 0.098(0.0437) (0.7330) (0.0854) (1228) (0.0303) (0.0103) (0.1070)

Social bonus (TS) -0.000 0.087 0.013 -448.5 -0.002 0.008 0.022(0.0300) (0.6290) (0.0871) (1210) (0.0334) (0.0122) (0.1240)

CrO visit (CrO) - 3.625*** 0.417*** 4997*** 0.006 0.015 0.426***(0.5030) (0.0921) (1257) (0.0302) (0.0112) (0.0955)

CrO * TC - -0.288 -0.018 -448.3 -0.011 -0.015 -0.084(0.7960) (0.1180) (1578) (0.0303) (0.0110) (0.1150)

CrO * TS - -1.158* -0.179 -1656 -0.015 -0.014 -0.190(0.6230) (0.1230) (1719) (0.0392) (0.0130) (0.1220)

P-value of F-test:TC = TS 0.388 - - - - - -CrO*TC = CrO*TS - 0.218 0.103 0.343 0.855 0.914 0.243

Observations 53,127 53,127 53,127 53,127 53,127 53,127 53,127No. of COs 4,380 4,380 4,380 4,380 4,380 4,380 4,380R-squared 0.089 0.069 0.013 0.014 0.041 0.048 0.039Mean dep. var., control 0.105 5.998 0.558 7022 0.832 0.966 -0.0723

Notes:The above regressions control for region and month dummies. The standard errors are reported in parentheses and are adjusted forwithin-field unit correlation between FAs; *p<0.1, **p<0.05, ***p<0.01.

34

C Appendix: Level of CO co-management

Figure C.1: Distribution of the share of CO-portfolio co-managed with other FAs

Note: The figure depicts the distribution of the share of FA’s CO-portfolio in the 9 months before thebonus was announced (June 2004 - March 2005) that was co-managed with other FAs. The dottedline in the graph shows the median value of co-management (73 percent of FA’s CO-portfolio). FAswith their share greater than the median value is categorized as “partnered” FA in the analysis.71 out of 81 partnered FAs co-manage their COs with one other FA, while the rest (10 partneredFAs) co-manage with two other FAs. FA partnership-teams are stable throughout the study period(unless one of the partners quit NRSP).

35

D Appendix: Balance tests on restricted samples

Table D.1: Summary statistics and balance tests (restricted sample, verified MPRs)


Demographic characteristicsAge 28.04 27.85 27.82 0.907 0.892 0.987Female 0.239 0.206 0.098 0.713 0.058 0.128Married 0.370 0.382 0.431 0.926 0.583 0.696Household head 0.174 0.176 0.176 0.970 0.974 1.000Completed high school 0.587 0.500 0.569 0.452 0.837 0.549Household consumption (Rs.) 7038 5927 6651 0.229 0.601 0.334Owns land 0.522 0.471 0.549 0.735 0.851 0.595Housing quality index 0.095 -0.077 0.349 0.361 0.222 0.074


Partnered FAa 0.522 0.412 0.608 0.441 0.588 0.208Overtime work (hrs/day) 2.145 1.833 2.278 0.360 0.608 0.191Time spent on microcredit tasks (%) 0.725 0.753 0.831 0.745 0.087 0.279Prefer credit bonus 0.565 0.500 0.549 0.596 0.898 0.711Social mobilization helps microcredit 0.957 0.941 0.941 0.708 0.732 1.000Did volunteer work before NRSP 0.326 0.206 0.333 0.238 0.945 0.279Best about NRSP: ability to help 0.413 0.529 0.608 0.316 0.106 0.471

Monthly performanceNumber of active loans 86.51 105.6 122.9 0.526 0.188 0.463New disbursement (Rs.) 65767 105845 107785 0.123 0.036 0.936Repayment on dues at 20th of month 0.746 0.718 0.668 0.654 0.225 0.545Repayment on dues at end of month 0.986 0.966 0.993 0.553 0.450 0.389Number of field units (FUs) 10 9 14 - - -Number of field assistants (FAs) 46 34 51 - - -Number of credit organizations (COs) 764 792 1150 - - -F-test statistics - - - 1.119 1.014 1.146P-value - - - 0.356 0.460 0.328

Notes: The standard errors (in F-tests in Columns 4-6) are adjusted for within-field unit correlation between FAs.aPartnered FA is defined as a dummy variable which equals one if an FA is co-managing more than 73 percent of her/hisCO portfolio (median value of co-sharing) with other FAs during the 9 months prior to the bonus announcement.

36

Table D.2: Summary statistics and balance tests (restricted sample, follow up)




Partnered FAa 0.519 0.414 0.531 0.513 0.943 0.487Overtime work(hrs/day) 1.914 1.805 2.233 0.810 0.381 0.277Time spent on microcredit tasks (%) 0.718 0.785 0.827 0.361 0.049 0.531Prefer credit bonus 0.630 0.483 0.551 0.292 0.543 0.646Social mobilization helps microcredit 0.926 0.931 0.959 0.909 0.402 0.520Did volunteer work before NRSP 0.241 0.207 0.327 0.775 0.414 0.361Best about NRSP: ability to help 0.519 0.448 0.531 0.443 0.908 0.492



37

Table D.3: Summary statistics and balance tests (restricted sample, post-bonus FA-CO)




Partnered FAa 0.509 0.387 0.519 0.427 0.951 0.430Overtime work (hrs/day) 1.897 1.884 2.266 0.975 0.313 0.301Time spent on microcredit tasks (%) 0.728 0.788 0.813 0.349 0.093 0.713Prefer credit bonus 0.623 0.516 0.538 0.453 0.539 0.876Social mobilization helps microcredit 0.925 0.935 0.962 0.807 0.354 0.533Did volunteer work before NRSP 0.245 0.194 0.327 0.653 0.463 0.310Best about NRSP: ability to help 0.509 0.484 0.558 0.789 0.653 0.548



38

Documents

Pay and Mission: The Impacts of Performance Incentives in ...economics.virginia.edu/sites/economics.virginia.edu/files/public/Gine.pdf · of group-based performance incentives for