T06P02 - Policy Implementation - The Role of Policy Targets

Topic : Policy Implementation

Chair : Anat Gofen - anat.gofen@mail.huji.ac.il

Second Chair : Robert Kent Weaver - Weaverrk@georgetown.edu

General Objectives, Research Questions and Scientific Relevance

Call for papers

Session 1 Firms and Institutions as Policy Targets

Thursday, June 29th 08:15 to 10:15 (Block B 5 - 1)


Robert Kent Weaver - Weaverrk@georgetown.edu - Georgetown University - United States

Anat Gofen - anat.gofen@mail.huji.ac.il - Hebrew University - Israel


Targets Heterogeneity, Ambiguity-Conflict and Policy Implementation: The Effect of Ownership on China’s Corporate Employee Pension Policy Implementation

Guo Lei - 06064@tongji.edu.cn - The department of public administration in the School of Economics and Management at Tongji University - China

Download the paper

This paper tests how heterogeneity in target groups affects China’s Corporate Employee Pension (CEP) policies implementation through the perception of ambiguity and conflict of the policy.


Although the policy implementation literature has indicated that target groups shape implementation greatly, how and to what extent target groups affect implementation have received limited attention. The present study proposes that the perceived policy ambiguity and conflict vary with heterogeneity in target groups, and then different implementation outcomes occur.


This paper puts the question into the context of China’s CEP policy, which consists of two parts: one is the mandatory Basic Pension (BP) policy, and the other is the non-mandatory Enterprise Annuity (EA) policy. Great implementation gaps have long existed. For BP policy, the real contribution rate is less than the nominal rate; for EA policy, the participation rate is extremely low.


The sample comprises all firms listed in China’s A-Stock Market from 2007 to 2014. The dependent variables are the real contribution rate for BP and participation for EA. The main explanatory variable is ownership, which segments enterprises into state-owned enterprises (SOEs) and non-state-owned enterprises (nSOEs). The two mediators, perceived ambiguity and conflict, are measured by housing benefits (real contribution rate for the housing benefits program) and average pay respectively. Path analysis is employed to estimate the direct and indirect effects of ownership on policy implementation. When the dependent variable is EA participation, Event History Analysis is applied, namely once a firm participated in EA, it is no longer in the dataset.


The results show that targets heterogeneity affects policy implementation directly as well as through the perception of ambiguity and conflict of the policy. The indirect effects work through 3 paths:

(1) heterogeneity-ambiguity-implementation;

(2) heterogeneity-conflict-implementation;

(3) heterogeneity-ambiguity-conflict-implementation.

The direct effect takes up the majority of the total effects for EA policy while the indirect effect does for BP policy. Policy attributes difference, mandatory vs. non-mandatory, maybe accounts for that.


This is relevant because the study develops a framework explaining how policy targets affect implementation and provides evidence with it. The paper also finds that the mediating effect of ambiguity-conflict is conditional on policy attributes.


Between Technical Expertise and Wise Counseling: The Role of Law Firms in the Implementation of Anti-Corruption Norms in Singapore

Alain Eloka - alain.eloka@unil.ch - Université de Lausanne - Switzerland

Download the paper

The role of private actors in the production and the implementation of collective norms has grown to an important issue of social-scientific inquiries over the last three decades. This has notably been investigated trough the analysis of Public-Private-Partnerships (PPP). While some scholars have found that private actors play a steady-increasing role in the production and the implementation of collective norms, others in turn argue that the state is still the key actor of policy processes. Meanwhile, no attention has been paid to the role of private actors in places where the state is still the major actor of social regulation. In this regard, observations about the role of law firms in the implementation of anti-corruption norms helps to shed some light in this blind spot of the literature. Drawing on empirical observations made in the process of anti-corruption in Singapore, this paper raises a major hypothesis about law firms and the implementation of collective regulatory norms, namely that these private actors play a key role regarding whether and how some policy targets such as business organizations comply or not with national and international anticorruption norms. It is thus a new input for scientific investigations into this domain, which in turn will contribute to a better understanding of the role of private actors in the making and implementation of public policies.

Keywords: Anti-Corruption policies – Collective norms – Implementation – Law Firms - Singapore


Policy targets' compliance with voluntary agreements: a different story?

Simona Torotcoi - torotcoi_simona@phd.ceu.edu - Central European University (Budapest, Hungary) - Romania

Download the paper

There always has been an interest in exploring compliance and implementation of international treaties and laws, from either the perspective of the involved actors, the process itself and its analysis, trends in development, or the issues the key stakeholders are dealing with. However, not until recently the topic of examining the implementation of voluntary policy agreements has started to be a topic of discussion. Enhancing policies that act in ways that are consistent with the goals and objectives intended by the policy makers in international voluntary agreements, represents a starting point for the members of that agreement to express their commitment.

Policy implementation scholars have offered several explanations for why and how a policy is implemented by putting as main variables the salience of the policy, its design and governance system, but also the arrangements and resources needed for operating that policy, including the will of the responsible bodies. Currently, there is no generally agreed theory of implementation but a collection of multi-factors influencing implementation. Implementation studies often show that targets' implementation actions either differ from the initially stated policy or these targets no not comply.

By using the case of higher education policy in the European Higher Education Area, and by combining the policy change and implementation literature this paper aims to shed light on the role of policy targets (university leaders, administrators, teachers and students alike) in policy implementation. First, it provides a conceptual clarification with regards to policy compliance and policy implementation. A framework which allows to dissect implementation in different stages (adoption, transposition, administrative and institutional implementation), and therefore, distinguish between different types and groups of policy targets, will be the base for conceptualizing and measuring different levels of compliance. In a second stage, by using the case of Romanian universities, this paper aims to support the proposed framework and to show how these levels are reflected at the national and institutional/university level. Last, the paper reflects upon the role of policy targets and their sources and usages of discretionary power, and extent to which the interaction between the lowest rank participants and the users transform not only the professional practices but also the public policies themselves.


Missing the target in Swiss “new regional policy”: Which factors help to explain the difficulty of addressing private actors as policy target groups?

Stefan Wittwer - stefan.wittwer@kpm.unibe.ch - KPM Center for Public Management at the University of Bern - Switzerland

Fritz Sager - fritz.sager@kpm.unibe.ch - KPM Center for Public Management at the University of Bern - Switzerland

Download the paper

This paper examines the implementation of a reform in Swiss regional policy in 2008, the “Neue Regionalpolitik” (NRP), on the basis of the conceptual implementation framework by Sabatier and Mazmanian (1980). The implementation of the distributional federal program takes place at the cantonal level and hence resulted in varying outputs of the policy. The aim of the NRP is to help rural, mountainous and border areas to implement their development programs in order to create and keep jobs in the areas by a direct financial promotion of initiatives, projects and programs. One specific and crucial policy output goal is the promotion of initiatives by regional entrepreneurs. Hence, regional entrepreneurs are the main policy target group aimed at. However, as Sager and Huegli (2013) pointed out in a comprehensive evaluation, the majority of the supported initiatives have been launched by public entities or associations (74%). The goal of this paper thus is to examine which factors lead to the difficulty of addressing regional entrepreneurs and hence to the problems of achieving a crucial policy goal. It is of high relevance at both the theoretical and practical level to understand why the policy failed to address its main policy target.

The framework by Sabatier and Mazmanian (1980) identifies factors which affect the achievement of statutory objectives (such as the policy output to address regional entrepreneurs) throughout the entire process of implementation and distinguish three categories: (1) The tractability of the problem(s) being addressed by a policy, (2) the ability of the policy to favourably structure the implementation process and (3) non-statutory (“political”) variables affecting implementation. With the help of data gathered in interviews with each implementation officer at the cantonal levels in the course of the evaluation by Sager and Huegli (2013), the three categories of the implementation framework can be operationalized regarding the cantonal level of implementation. Additionally, data at the project level is available for 1248 projects (collected by CHMOS).

To test the implementation framework and examine whether the three categories explain the difficulty in addressing the preferred policy target, bayesian multilevel logistic regressions will be conducted, testing what contributes to the explanation whether a project is launched by an enterprise or by public entities and associations. On Level 1, project specific aspects such as the project type (e.g. tourism), costs or instrument of support are being considered. Level 2 addresses cantonal characteristics regarding the implementation framework such as policy inheritance ((1) problem tractability), size and willingness of bureaucracy ((2) Structuring by the policy) and (3) non-statutory factors such as population density or political factors.

Preliminary results indicate that factors on both levels contribute significantly to the explanation of the policy output whether regional entrepreneurs are addressed as policy target groups and hence allow to assess the conceptual implementation framework by Sabatier and Mazmanian (1980) thirty-six years later.

Session 2 Governments and Bureaucrats as Targets

Thursday, June 29th 10:30 to 12:30 (Block B 5 - 1)


Anat Gofen - anat.gofen@mail.huji.ac.il - Hebrew University - Israel

Robert Kent Weaver - Weaverrk@georgetown.edu - Georgetown University - United States


Policy noncompliance and policy change – the case of local government amalgamation

Oddbjørn Bukve - oddbjorn.bukve@hvl.no - Western Norway University of Applied Sciences - Norway

Download the paper

This paper studies local government amalgamation reform in Norway in order to analyse policy noncompliance by policy targets, together with the government’s response to noncompliance (Gofen, 2015). By focusing noncompliance as a strategy to avoid or change a policy, we study target resistance and government reactions as an interactive process characterized by policy dissonance (ibid.), where noncompliance may influence policy change over time. Gofen (ibid.) distinguishes between enforcement, accept, adaptation and embracement as government responses to reconcile policy dissonance. Through case studies based on theoretical sampling, Gofen has shown how these concepts are fruitful for analysing responses to policy dissonance. I deploy her conceptual framework to analyse a different kind of policy targets – local governments, and a complex reform policy where a number of governance tools are in use.


Overall, noncompliance in the amalgamation reform is high. Of 428 local governments, only 88 have decided to merger into 36 new LG’s. 55 additional LG`s want to merge, but have not found willing partners. Regarding types of noncompliance, we find both rejection of specific reform elements and broader, value based voices. Main examples of the first type is noncompliance because local actors do not see a need for policy change, together with fear of losing nearness to LG services and institutions. Value based noncompliance is found where local actors maintain that amalgamation will reduce local democracy, and where they resist the reform because it is viewed as an element in a broader centralization strategy from the government. 


When classifying policy elements and responses I use the NATO typology (Hood & Margetts, 2007) of policy tools in combination with Gofen’s response typology. Regarding nodality, initial policy established clear criteria for municipal size. Later responses modified the ambitions, pointing towards acceptance of noncompliance. Regarding treasure, there is a development from incentives to sanctions, pointing towards enforcement. There is mandatory local negotiations and an organisational setup for process advice from county governors. The advisory process varies, from embracement to enforcement. Authoritative decisions are to be made by theparliament in the spring 2017, hence the response is not known yet.


The main conclusion is that government responses to policy dissonance are not consistent so far. The government has to handle different target groups who do not give the same reasons for noncompliance, and this make it difficult to establish consistent responses.



Gofen, A. (2015). Reconciling policy dissonance: patterns of governmental response to policy noncompliance. Policy Sciences, 48(1), 3-24. doi:10.1007/s11077-014-9202-9

Hood, C. C., & Margetts, H. Z. (2007). The tools of government in the digital age (2nd rev. ed. ed.). Basingstoke: Palgrave Macmillan.


Performance Measurement, Policy Compliance and Noncompliant Behavior: The “Fatality Quotas” in China’s Work Safety Management

JIE GAO - polgj@nus.edu.sg - National University of Singapore - Singapore

China has been accused of achieving its impressive reform-era GDP growth at the expense of numerous human lives lost in work-related accidents. The so-called “blood-soaked GDP” not only provides fodder for China’s critics in the international community, but also jeopardizes social stability domestically. To tackle this issue, in 2004, the Chinese central authorities established a system of “fatality quotas”—fatality numbers and ratios that must not be exceeded in various types of work-related accidents—in the hope of reducing the loss of lives. Since then, the central authorities have prioritized these fatality quotas in the format of performance targets in the evaluation of local officials’ work accomplishment. Local officials are held strictly accountable for their performance in staying within the indicators’ parameters, and their careers will be jeopardized if they exceed the death indicators’ limits.


Why did the Chinese leaders decide to adopt the fatality quota system—apparently a weirdly designed tool—to deal China’s work safety problem? How effective has this system been in ensuring local bureaucrats’ compliance with the national goal of reducing China’s work-related fatalities? What would happen if it were impossible for local officials to comply with the allocated quotas? This study aims to answer these questions. It adopts a qualitative method. Empirical data are mainly drawn from national and local party documents (many are circulated internally within local governments) and the author’s fieldwork in several Chinese localities in the past ten years.


This study contributes to the discussion of this panel in two major aspects: First, it provides a concrete case study—and a remarkable example—to illustrate why and how policy targets are used to strengthen local bureaucrats’ compliance with the national development goals, especially in a transitional society. Second, it examines the noncompliant behaviors caused by the implementation of such policy targets, and joins the discussion on how or whether noncompliant behavior might be prevented in the policy design process.


Gold Monetization in India as a Transformative Policy: A Mixed Method Analysis

Priya Narayanan - priyan@iima.ac.in - Indian Institute of Management, Ahmedabad - India

Balagopal Gopalakrishnan - balagopalg@iima.ac.in - Indian Institute of Management, Ahmedabad - India

Arvind Sahay - asahay@iima.ac.in - Indian Institute of Management - India

Download the paper

India is the second largest consumer of gold in the world and gold is a major contributor to the current account deficit. Much of the gold goes out of circulation and is not available to support economic activity. To encourage consumers to bring the gold back into circulation, the government of India instituted the Gold Monetization Policy in 2015. This research views the Gold Monetization Policy in India through the lens of consumer associations with gold, as well as the banker and refiner perspectives on implementation challenges. The success of this policy is important for the country to better manage its current account balance, in a milieu where gold consumption holds sociocultural importance. The study uses an empirical approach to analyse how various policy targets have approached the policy, and provides suggestions to increase compliance.


First, a nationwide survey-based study of 1171 households, across 10 states that constitute approximately three-quarters of annual national gold consumption, was conducted to understand the consumer associations with and attributions related to gold. This shows family functions and festivals to be triggers for gold purchase, indicating ingrained habit and planned accumulation.  There is also high liquidity and safety association of gold, which is also not considered as having any substitute, along with a clear reluctance to sell gold received as a gift. Rural consumers are more reluctant to part with gold as compared to urban consumers, but are also ready to pledge gold as collateral, suggesting “liquidity use” of gold on the basis of requirement.


Second, an interview based study was conducted with senior management of 6 banks, 5 refiners and one industry consultant to understand the challenges and implications of the policy for targets. Banks would promote products based on this policy if they had more control and if there was clear separation or effective mitigation of risks relating to the operationalization of the policy. Lack of sufficient incentive alignment is the main reason why banks are not yet fully on board. Finally, an econometric analysis of gold consumption and its potential determinants was conducted using household data from all 640 districts of the National Sample Survey for 2011-12. The analysis shows that propensity to consume gold is positively correlated with proportion of females in the household and with number of daughters in the household. Also, households in districts with higher sex ratio have a higher propensity to consume gold, rural households have a higher propensity to consume gold and Hindu households have a higher propensity to consume gold. Based on the research, several steps are proposed to enhance the adoption of this policy by all targets.


The research considers targets’ compliance/noncompliance as the dependent variable, and uses a mixed method approach to understand the targets’ motivations and barriers to compliance.

Session 3 Individuals as Policy Targets

Thursday, June 29th 13:30 to 15:30 (Block B 5 - 1)


Fritz Sager - fritz.sager@kpm.unibe.ch - KPM Center for Public Management at the University of Bern - Switzerland

Anat Gofen - anat.gofen@mail.huji.ac.il - Hebrew University - Israel


Getting the full picture: an empirical framework for understanding gaps in targets' behaviors.

Olejniczak Karol - k.olejniczak@uw.edu.pl - EUROREG, University of Warsaw - Poland

Pawel Sliwowski - p.sliwowski@uw.edu.pl - University of Warsaw, The Centre for European Regional and Local Studies - Poland

Magdalena Roszczynska-Kurasinska - magda.roszczynska@gmail.com - Uniwersity of Warsaw - Poland

Download the paper


Public policies aim to guide citizens towards socially desirable behaviors (Datta & Mullainathan, 2012; Shafir, 2013a). During policy implementation real behaviors of policy targets (e.g. citizens, private or public organizations) often differ from those expected by policy makers. Yet, for a long time (non)compliance was not systematically addressed in policy discourse neither from theoretical nor practical perspective (Weaver 2009, Etienne 2010).


As shown by Etienne (2010) there have been three major approaches for explaining this issue: neo-institutionalism, social norm perspective, and realistic models of behaviors. Additionally, recent theoretical and practical developments (Weaver, 2015; Hill & Hupe, 2014, Stanford University, 2016) offer promising approach to enrich our understanding of non(compliance). They focus on barriers and gaps that hinder targets' compliance. The noncompliance is attributed either to gaps in motivation or targets' capacities.


Building on that developments we argue that behaviors of policy targets are results of complex mechanisms driven by: characteristics of individuals, contextual constrains and target's motivations (Michie, van Stralen, & West, 2011; Ostrom, 2010; Astbury & Leeuw, 2010; Pawson, 2013; World Bank, 2015). There is a need for a holistic view on those mechanisms. The fragmented view of target's gaps at the stage of policy design may result in policy failures at the stage of implementation.



We are, thus, motivated to ask: ”How can we conceptualize targets' behaviors in order to systematically describe and effectively address their different dimensions during policy design?


This paper offers an analytical framework for holistic mapping of gaps and barriers in mechanisms of policy targets’ behaviors. The discussion is illustrated with an empirical case study of policy implemented in 2015 by the government of Poland to address the obesity problem of school children.



Our paper is divided into three parts. First part provides a conceptual framework for mapping barriers and gaps in targets' behaviors. It adapts COM-B model that defines behavior of policy targets as a combination of capacities, opportunities and motivations (Michie et al., 2011; Michie, Atkins, & West, 2014). This frame is complemented with decision-making heuristics (Shafir, 2013b), resources and rules (Ostrom, 2010), and barriers to behaviors (Stanford University, 2016; Weaver, 2015).


In the second part, we use this conceptual framework to explain discrepancy between policy design and policy implementation. We study case of a policy introduced in Poland to reduce child obesity. The effects of this policy have been explored with desk research of legal documents, systematic literature review on interventions targeting obesity of pupils, and ethnographic multi-stakeholder research in selected Polish schools.


In conclusions we discuss strengths and weaknesses of the proposed framework as well as implication for policy practice.



Paper contributes to the conceptualization of determinants of policy compliance by proposing a coherent framework and a set of definitions. It also helps practitioners to think systematically about gaps in targets' compliance when designing policy solution. That could prevent policy designers from overlooking real drivers of targets' behaviors, and it could increase the chances of policy implementations.



Designing a Performance Measurement System in case of Time Bound Service Delivery Act (SAKALA) in Karnataka

Vishwanathan Iyer - viyer@tapmi.edu.in - T.A Pai Management Institute - India

Varsha Khandker - varsha@tapmi.edu.in - T.A Pai Management Institute - India

Kedar Joshi - kedarj@tapmi.edu.in - T A Pai Management Institute - India

Vidya Pratap - vidyapratap62@gmail.com - T. A. Pai Management Institute - India

Download the paper

Designing a Performance Measurement System in case of Time Bound Service Delivery Act (SAKALA) in Karnataka


Vishwanathan Iyer, Dayashankar Maurya, Kedar Joshi, Varsha Khandkar & Vidya Pratap

T A Pai Management Institute, Manipal, India


Extensive research has been devoted to identify the risks imposed by performance measurement system, and mitigating strategies for the same for designing an ‘authentic performance’ measurement system. As most of these studies are conducted in developed country context, we know very little about the extent to which these learnings have been transferred to public administrative reforms undergoing in developing countries. Under new public management reforms, performance measurement in developing countries are getting reformed aided by the growing use of IT enabled tools to measure and manage performance, but there have been limited studies examining design and management of these performance measurement system.


Under the SAKALA Act, the Karnataka government (India) is committed to provide services identified under the act in a time bound manner. Using this natural setting, we analyze design and management of performance measurement system for 776 public services in the Udupi district (one of districts under the state of Karnataka). We compare the performance measured under existing performance measurement system vis-a-vis performance measured using the approach advocated on the basis of an 'authentic performance measurement system'. The Analysis is based on three year administrative data on performance indicators for 776 services, in-depth interviews and analysis of official documents.


The comparative analysis reveals that present performance analysis and measurement approach lacks accuracy, doesn’t pinpoint low performing services and departments, and mis-represents performance by wrongly aggregating performance to create a favorable impression. Thus the present performance analysis rules out opportunities to identify what contributes to poor performance, critical for any effort to diagnose it.  The present performance measurement system also doesn’t tie performance to actors that ‘make up’ performance but instead punishes front line workers not responsible for poor performance. Thus the existing performance measurement doesn’t enable to identify weak links in service delivery, doesn’t identify people and reasons for poor performance and therefore provides little guidance on how performance can be improved. Hence present performance measurement and management system do not perforce performance the sine qua non of ‘authentic’ performance management systems. The case clearly illustrates that technology is only an enabler and there is little transfer of learning in designing and management of performance measurement systems.  


Key Words: Performance Measurement, Performance Management, Time Bound Service Delivery, Public Services, SAKALA


Securing Compliance: the Collective Aspects

Drorit Gassner - drorit.gassner@mail.huji.ac.il - The Hebrew University of Jerusalem - Israel

Download the paper

Securing Compliance: the Collective Aspects

Drorit Gassner


Hebrew University

Paper proposal for ICPP 3, Singapore 2017


Compliance of policy targets' with policy is essential to successful implementation. Therefore, noncompliance is often approached as an implementation problem followed by increased governmental enforcement efforts. Although governmental enforcement efforts vary, it mainly involve incentives and information in an attempt to bring policy targets’ behavior into line with current policy arrangements. Following deterrence’s limitations indications, greater emphasis is given recently to responsiveness, flexibility, and creativity as key components in enforcement, which is expected to increase compliance by encouraging policy targets’ collaboration, cooperation, and engagement in the provision of public services. A more preventive approach suggests moving from responding to noncompliance after implementation, to bearing in mind targets' capacity to comply already during the policy design stage. Importantly, compliance as well as noncompliance are considered individual's behavior and accordingly, policy targets are approached as a collection of individual policy-clients rather as a collectivity. To focus on collective aspects of compliance, this study focuses on those who manage public service provision to a defined locality in three different policy sectors: policing, education and welfare. Analysis draws on in-depth semi-structured interviews (N=78) with police station chiefs, primary school principals, and heads of social services bureaus in Israeli local authorities. Preliminary findings suggest that collective compliance efforts are ongoing and draw on the establishment and maintenance of continuous relationship with locality leaders. Specifically, findings indicate three roles through which compliance is secured, namely, negotiation mediation, and entrepreneurships. Negotiation is manifested through assimilation of locality needs and requirements during the translation of formal policy directives to daily work routines. Mediation refers to the need to balance and resolve conflicting policy needs of different sub-groups within the locality. Entrepreneurship refers to the ways locality is mobilized to engage in service provision. Focusing on policy clients as collectivity allows a more nuanced understanding of securing compliance and its contribution to implementation success.        


The Government Response to Noncompliance and its Limitation on Primary and Secondary Education in the Lao PDR

Atsuo SATO - atsuo@soleil.ocn.ne.jp - The University of Tokyo - Japan

Download the paper

In many developing countries, behavioral change of policy targets has occurred dramatically and collectively in the process of working towards the United Nations Millennium Development Goals (MDGs). Lao PDR, a Southeast Asian country, has made great progress in education and gender equality. The percentage of net enrollment in primary education increased from 58.8% in 1992, to 98.5% in 2014. The ratio of girls to boys in primary education improved from 0.79 in 1990, to 0.91 in 2012. These outcomes result not only from the institutional building on compulsory education and the capacity building of teachers, but also from the governmental response to noncompliance. First, this study examines the mechanisms to increase compliance in primary and secondary education in the surveyed villages in Lao PDR. Several government responses to barriers to behavioral change were identified through focus group discussion and in-depth interviews. Second, the study focuses on the role of a women’s organization and its network, which provide information and persuade village women (mothers of children) to comply with government policies. The women’s organization and its network work well in the villages where the majority of the population is of the Lao ethnic group. However, they do not work in the villages where mostly Hmong people (one of 49 ethnic groups in Lao PDR) live. The difference in performance is explained by their kinship structure and their history of state building. Third, the study examines the reasons for the noncompliant behavior of Hmong girls in secondary education. Primary education for girls has already been accepted as a social norm and a family value. Villagers, however, have only partially accepted the importance of secondary education for girls. Noncompliant behavior in secondary education results from more fundamental life circumstances: The Hmong live in a highland area and rely on subsistence farming. Noncompliance is connected with their safety net, because getting married and giving birth early in life means their child can take care of them from an early stage. Therefore, girls often drop out of secondary school at around 15 years of age to get married. Through case analyses of two ethnic groups, this study offers a nuanced understanding of compliant and noncompliant behavior in a developing country.

Export PDF