Devolution of Power and Public Policy: The Potential of Local - - PDF document

Devolution of Power and Public Policy: The Potential of Local Governments to Address Population Issues

Paper presented at the IUSSP 2017 International Population Conference, Cape Town, 30 October 2017; Session 1901—Population and policy changes in Africa AUTHORS Beth Fredrick1,2 Celia Karp1 AFFILIATIONS

1Department of Population, Family and Reproductive Health, Johns Hopkins Bloomberg

School of Public Health, Baltimore, MD, USA; 2Bill and Melinda Gates Institute for Population and Reproductive Health, Baltimore, MD, USA CORRESPONDING AUTHOR Beth Fredrick Department of Population, Family and Reproductive Health Johns Hopkins Bloomberg School of Public Health 615 N. Wolfe Street Baltimore, MD 21205 bethfred@jhu.edu

1 ABSTRACT Background: Devolved governance, the transition of power from national to subnational government units (SNGs), is nearly universal in sub-Saharan Africa. Devolution, sometimes referred to as decentralization, encompasses legal, political and fiscal control, administration, and service provision, which support local priority-setting, decision- making, and investments. Despite increased decentralization, little is known about how it has affected development and implementation of policies related to population issues at the subnational level. Understanding the parameters of subnational government decision- making and the evidence used to inform these decisions is critical to achieving health and development goals. Methods: A literature review was implemented to identify information on decentralization and data for decision-making in sub-Saharan Africa. Key informant interviews were conducted with public health practitioners at the international, national, and subnational levels to explore approaches to evidence-based subnational decision-

• making. Analysis was restricted to demographic and health information sources related to

family planning in sub-Saharan Africa. Results: More than 40 articles were reviewed and 25 informants were interviewed. Limited data exist on the financial, political, and service provision authority of SNGs in

• Africa. Informants emphasized utility of five data systems within sub-Saharan Africa

(DHS, DHIS2, HMIS, MICS, and PMA2020), stressing the limitations of these systems for subnational decision-making. Subnational reliance on antiquated data systems coupled with data literacy challenges among local leaders restrains the translation of data into practice. Discussion: The unique needs of SNGs for reliable and useful data to support national and global goals are not being met. Improved availability and use of data within decentralized countries could support SNGs to propose, enact or implement productive changes in policy and programs addressing population health needs. The responsibility is two-fold: SNGs should better inform data priorities, collection, and interpretation; researchers, analysts, and those supporting data collection and dissemination should investigate and anticipate needs for accessible and relevant subnational data within specific geographies.

2 INTRODUCTION Devolved governance, the transition of power from national to subnational government units (SNGs), is increasingly common worldwide and nearly universal in sub-Saharan Africa.1 Devolution, sometimes referred to as decentralization, encompasses legal, political and fiscal control, administration, and service provision, all of which support local priority-setting, decision-making, and investments.2 In conjunction with these authority changes, decentralization has diverse implications related to democratization, social, political, and economic development, information flow, and government transparency and accountability. At the core of decentralization is the concept that shifting decision-making closer to constituents themselves will result in improved policies and programs that reflect and respond to local priorities.3 In the era of Sustainable Development Goals (SDGs) and the Family Planning 2020 (FP2020) partnership, national commitments to improved family planning (FP) access have positioned country governments with a need for data measurement at national and subnational levels. Exploring decentralization and the influence it has on where, how and when data is collected, analyzed and interpreted is equally important to development and the fulfillment of these national commitments.4 Decentralization is directly relevant to the ability of countries to adapt to results-based financing mechanisms for reproductive, maternal, newborn, child and adolescent health (RMNCAH) such as the World Bank’s Global Financing Facility (GFF). The GFF “recognize[s] that countries themselves are the engines of progress and that the role of external assistance is to support countries both to get more results from the existing resources and to increase the total volume of financing”.5 One of the core tenets of the GFF is the strengthening of systems to track progress, learn, course-correct, and help countries gain financial support for their own priorities. Kenya is just one example of a decentralized country that has already incorporated county governments in its GFF investment framework.6 Other countries are likely to follow suit in the recent effort to engage and include subnational governments, if not in developing their investment case, then certainly in its implementation.

3 Despite rapid increase of decentralization since the 1990s and transfer of certain responsibilities to SNGs, little is known about how this process has affected development and implementation of financial and health policies related to population issues at the subnational level.7 In part, this is a consequence of the diverse characteristics of decentralization and the variety of forms it takes in different country contexts, but it is also related to the ever-changing dynamics between central and subnational governance, priorities, and demands.8 Regardless, subnational units are commonly responsible for planning, budgeting, and resource allocation and implementation of national policy. In order for subnational units (e.g., states, counties, provinces, districts, communes, municipalities) to fulfill these responsibilities and respond to local issues, they must first be equipped with the necessary tools to identify the issues in their own geographies and understand how to measure and monitor them. Beyond addressing the needs of their own constituents for contraceptive access, information, and quality services, understanding the parameters of SNG decision-making related to family planning and the evidence used to inform these decisions is critical to achieving global population health and development goals. The potential of subnational governments to contribute to global health goals and the constraints they face in effectively using data for decision-making are gaining

• prominence. In the words of Michael Bloomberg, United Nations (UN) Special Envoy for

Cities and Climate Change and three-term New York City Mayor, a key mission of support for subnational policymakers is to “create ways to measure whether we’re actually meeting the goals, because if you can’t measure it, you can’t manage it. Government programs never want measurement because then you have to implement it, you have to pay for it—they don’t want to do that. And if you implement it poorly, they don’t want to be held accountable.9” Conversely, as national and subnational leaders make financial decisions, design and manage programs, and monitor performance, effective collection and use of population data will drive innovation and economies of scale. As noted by experts convened by the

4 Sustainable Development Solutions Network in 2015, “encouraging new, reliable, and accessible government data, will provide governments the capacity to design better policies and programs. It will also enable citizens to hold leaders accountable for progress and improve their day-to-day decision-making”.10 Yet, even with this recognition, little of the growing attention paid to improving data for decision-making at the global and national level has acknowledged or adequately incorporated the needs of subnational leaders. This paper focuses on the potential of decentralized governance for addressing population issues, emphasizing decisions related to delivery of contraceptive information, services, and supplies in sub-Saharan Africa. Research Questions and Focus

• 1. ¡ To what extent do SNGs in sub-Saharan Africa use data to inform their decision-

making related to population, health, and development issues and what are the sources of those data?

• 2. ¡ How reliable, available, and useful are country-specific data systems for SNGs?
• 3. ¡ What ways are data limited or restrained in their capacity to inform subnational

decision-making related to population, health, and development issues?

• 4. ¡ What is feasible and most urgently needed to improve the ability of SNGs use

data effectively in meeting the needs of their communities and contribute to the achievement of national and global goals? To answer these questions, this research focuses on the following: 1) synthesis of related literature and contributions of public health practitioners at international, national, and subnational levels with diverse experience in data collection, analysis and application related to family planning and global compacts (i.e. the SDGs and FP2020) within sub- Saharan Africa; 2) identification of data deficits at subnational levels in sub-Saharan Africa for decision-making surrounding population, health, and development issues; 3) translation of results to inform recommendations for consideration by governments, donors, and researchers to address the SNG data needs.

5 This analysis is exploratory; it is grounded in the experience of the Advance Family Planning (AFP) evidence-based advocacy initiative within the Bill and Melinda Gates Institute of the Johns Hopkins Bloomberg School of Public Health and the authors’ involvement in that initiative. Since 2011, AFP has worked with hundreds of SNGs to improve access to and use of quality family planning services and to assess the impact of their decisions and actions.4 This paper serves as a first step toward gauging the scale of data use among SNGs in sub-Saharan Africa, specifically in SNG decisions related to family planning and strategies for FP data improvement in terms of availability, reliability and use. To set the context, this paper provides a brief rationale for and intentions of decentralization; the scope of devolved governance within sub-Saharan Africa; and the availability of and demand for evidence to inform subnational policy and program change. METHODS A literature review was conducted to identify the means through which devolved governance in sub-Saharan Africa shapes establishment and implementation of population, health, and development policy decisions. Relevant information sources on decentralization in sub-Saharan Africa and data systems that exist within local government contexts were identified through parallel searches in PubMed, Scopus, and ScienceDirect, using related search terms and inclusion criteria (Table 1).

Table 1. Literature Review Search and Inclusion Criteria Search Criteria Terms Africa, sub-Saharan, devolution, decentraliz*, decentralis*, subnational, district-level, governance, government, decision- making, policy, authorit*, population, health, develop* Databases PubMed, Scopus, Science Direct Inclusion Criteria

• 1. ¡ The article analyzed decentralization in sub-Saharan Africa.


• 2. ¡ The article assessed data used by subnational government units. 

• 3. ¡ The article examined population, health, and development issues.
• 4. ¡ The article was published in English. 

• 5. ¡ The article was published in a peer-reviewed journal. 

• 6. ¡ The article was published between January 1, 1990 and November 15, 2017. 


6 Based on the preliminary research, three core measures of data systems were established: availability, reliability, and utility of data.11 Availability was defined as existence and accessibility of subnational data; reliability as the quality, trustworthiness and acceptability of subnational data; and utility as application and interpretation of subnational data. These measures were used to guide the development of a structured interview guide for the key informant interviews (KIIs). This paper focuses on the qualitative component of the research, which aimed to understand availability, reliability, and utility of and need for evidence to inform subnational family planning policy and program change in sub-Saharan Africa. The authors conducted KIIs with 25 government leaders, global health implementing agencies, and other stakeholders from international and sub-Saharan African

• rganizations via Skype and in-person. Interviews were transcribed verbatim and

reviewed for common themes surrounding availability, reliability, and utility of subnational data systems. Direct recommendations from KIIs, as well as informal recommendations deduced from other responses throughout the interviews, guided recommendation development that constitutes the discussion section of this paper. The Institutional Review Board (IRB) at Johns Hopkins University Bloomberg School of public Health deemed this research IRB-exempt after review of the proposed research methods and scope of this analysis. RESULTS LITERATURE REVIEW The three database searches yielded 704 articles, with 226, 416, and 62 articles in PubMed, Scopus, and ScienceDirect respectively. Of these 704 articles, only 26 articles met the inclusion criteria. Additional resources, including existing analyses of devolution by the World Bank, United Nations, and Organisation for Economic Co-operation and Development and those recommended by key informants, were also included. Though not the focus of this paper, more than 40 relevant articles and reports ultimately informed the authors’ analysis of data application for health and development policy within SNGs.

7 Decentralization, Development and Data for Decision-making The World Bank was one of the initial leaders in the 1990s to highlight the complex ways that decentralization was and would continue to affect SNGs, including shifts in resource mobilization and allocation, demand for knowledge of local systems, and understanding

• f what approaches are effective.7 As the World Bank noted at the time, one expected
• utcome of decentralization is that improvements at the local-level accelerate national

development, including heightened development in rural areas.1 The expectation is built

• n the premise of country ownership, which has been defined in many ways over the past

two decades. One definition that is relevant to this inquiry and analysis refers to internal

• r democratic ownership, citing local and provincial governments as among the many

influential actors beyond national governments. This definition recognizes the need to build local capacity to implement and monitor development activities12; an aspiration that requires data availability, reliability, and utility to be realized. A prime example is the establishment of SDGs in 2015. The SDGs set ambitious benchmarks for 193 countries to improve their health systems and other development measures.13 Nine targets and related indicators that define SDG3, which focuses on health and well-being, enable institutions to track country-level progress through 2030. Researchers and others have highlighted the need for country-specific data to identify where countries should concentrate their investments, efforts, and resources.14 Acceleration of decentralization over the last three decades resulted in a dramatic growth in the number of subnational government units throughout sub-Saharan Africa (Table 2). With the establishment of each level of subnational governance, financial, administrative, legal, and service provision decisions were shifted downward and local priorities began to guide local policy development and implementation.15 As more and more SNGs were created, gaining decision-making power and control, capacity-building for effective governance management and development progress at the subnational level was required.16 Although the literature review revealed a dearth of research synthesizing the varying levels of authority, control, and responsibility throughout sub-Saharan Africa, it is evident that dispersion of power during this period transformed governance and

8 decision-making—the historical centralized approach took less precedence and local needs, interests, and priorities came into focus.

Table 2. Subnational government units (SNGs) in sub-Saharan African countries. Subnational government unit level Number of SNGs Regions/States/Provinces/Zones 1,500 Districts/Departments/Counties 7,150 Local administrative units 55,212 TOTAL 63,910 Estimates based on national government website data across all sub-Saharan African countries. Variance

• f these estimates is expected in geographies experiencing conflict, unrest, or major governance changes.

Local administrative units defined as any government unit lower than a district department, or county.

SYNTHESIS OF KEY INFORMANT INTERVIEWS Twenty-five key informants, representing local organizations in Uganda, Nigeria, Tanzania, and Ethiopia, as well as international NGOs in the United States, provided insight into data for subnational decision-making surrounding population, health and development issues in sub-Saharan Africa. While SNGs are increasingly relied upon to inform those decisions, three key factors found were identified as central to assessing and improving the capacity of SNGs to use data effectively to support health and development goals, such as those in the SDGs and the GFF. These are not limited to the sub-Saharan Africa political landscape, SNG decision-making, or family planning policy in particular, but they have implications for all three. First, there is limited information on the specific financial, political, and service provision authority of SNGs in Africa. To date, no research has synthesized the varying levels of SNG authority in sub-Saharan Africa to specify the scope of power and control that subnational government units have to address population, health, and development issues in their local contexts. Without the mandate to address needs directly, it is unlikely that these decision-makers will have the time or inclination to use data, regardless of whether it is available or reliable.

9 Second, there is limited information about the capacity of national and subnational data systems to provide trusted and relevant population, health, and development information, including the human resources and funding necessary for the effective operation of those

• systems. Many interviewees cited the absence of processes to enable SNGs or health
• fficials to suggest areas of inquiry, improvements in data collection, or possible errors

within extant data sources. Tailoring interventions and strategies to the strengths, needs, resources, and values of particular populations and localities coupled with systematic learning and documentation could inform future action.17 Third, the promise of the internet, geographic information systems (GIS) technology, and cellular connectivity is far from realized in sub-Saharan Africa and compounds the challenges of this area of inquiry. Even as the information ecosystem evolves, access to information in SNGs will be compromised by lack of access and connectivity in poor and rural areas, monopolization of information access at higher levels of authority, lack of mobile literacy, and restrictions on freedom of expression.18 As one informant noted, “sometimes lack of access to data is as simple as lack of electricity or paper on which to print a document or a slow internet connection”. Investments in Data for Decision-making—Implications for SNGs Sound and reliable information is the foundation of decision-making across all health system building blocks. It is essential for health system policy development and implementation, governance and regulation, health research, human resources development, health education and training, service delivery and financing. The health information systems have four key functions: data generation, compilation, analysis and synthesis, communication and use.19 Interviewees recognized the value of a range of information sources that may be used by SNGs. Nonetheless, they consistently underscored the need to better match data collection and dissemination efforts to the needs of SNGs seeking to improve the health and well-being of their localities.

10 The essential role of data in informing and guiding government policy and programs on reproductive health, including family planning, and development has a long history (Table 3). In 1972 the International Statistical Institute established the World Fertility Survey, which was succeeded by the Demographic and Health Surveys (DHS) in 1984.20-

21 The DHS program’s main objective is to improve and institutionalize the collection

and use of data by host countries for program monitoring and evaluation and for policy development decisions.22 With USAID support, the DHS has collected, analyzed, and disseminated accurate and representative data on population and health through more than 300 surveys in over 90 countries. Surveys of a nationally representative sample of women and men measure key indicators, including fertility and family planning use are typically conducted every five years. Respondents emphasized that Sub-Saharan African countries have come to rely on the DHS to mark progress and address fertility, family planning, and women’s health issues, among others. Countries recognize the need to more effectively respond to and monitor rapidly changing health needs and new developments. Examples of such developments in family planning include contraceptive stockouts or introduction of new technologies, such as subcutaneous Depo-Provera (DMPA). In tandem, increases in international funding for health and performance-based financing mechanisms such as the GFF have, fueled a demand for more reliable, regular and timely data.23 Donors, such as USAID and NORAD have invested in improvements in national health management systems (HMIS) and district health information systems (DHIS and DHIS2). Concurrently, the need for more timely and targeted information has supported development of complementary efforts, including the Multiple Indicator Cluster Surveys (MICS) supported by UNICEF and the Performance Monitoring and Accountability 2020 (PMA2020) initiative of the Bill and Melinda Gates Institute for Population and Reproductive Health at the Johns Hopkins Bloomberg School of Public Health.

11

Table 3. Data systems collecting reproductive health information at the subnational level. Mechanism Description Strengths Limitations Demographic Health Surveys (DHS) World’s largest, nationally representative household surveys collected in 90+ countries since 1984. Focus

• n reproductive and child
• health. Led by USAID.
• ¡ High response rates,

national coverage, high- quality modular data

• ¡ Comparability with

MICS

• ¡ Administered at

different times in different countries

• ¡ Limited subnational data

District Health Information Software (DHIS2) Software platform for health systems data collection utilized in 40+ countries for health metrics reporting, analysis and dissemination.

• ¡ Captures aggregated and

health event data

• ¡ Widely used at the

subnational level

• ¡ Limited to data in

health facilities

• ¡ High level of error and

missingness in the data Health Management Information Systems (HMIS) Country-specific data systems capturing publicly provided health services and health outcomes data.

• ¡ Public facilities report

into system on continuous basis

• ¡Excludes relevant data

from private facilities

• ¡Quality varies greatly

by country Multiple Indicator Cluster Survey (MICS) Nationally representative household surveys collected in 100+ countries since

• 1995. Focus on children.

Led by UNICEF.

• ¡ High response rates,

national coverage, high- quality modular data

• ¡ Comparability with DHS
• ¡Two-year reference

period for recall

• ¡Limited SRH data
• ¡Limited subnational data

Performance Monitoring and Accountability 2020 (PMA2020) Mobile technology platform for routine data collection

• n family planning

indicators, and other health measures, in 10 countries.

• ¡ Rapid turnaround from

data collection to analysis

• ¡ Household and facility-

level data obtained

• ¡Only available in

countries where PMA2020 works

• ¡Subnational data do not

produce reliable estimates in most PMA2020 countries

Few of these data systems adequately address the needs of SNGs for a variety of reasons. According to key informants, inconsistencies, incomparability, and inadequate sample sizes for subgroups and at subnational levels hampers the ability of health officials to use data effectively. Furthermore, sub-Saharan African SNGs in particular are not routinely consulted on their data needs or supported in data assessment, analysis or dissemination to their colleagues and communities. Finally, with regard to family planning, SNGs need more specificity related to contraceptive method availability and use, the accessibility and quality of services, unintended pregnancy and the consequences, and inequities (e.g. whether young people are able to access and use contraception) within their own locales.

12 Recent Initiatives and Investments in Data for Decision-making Recently, a range of global efforts are making progress in addressing the need for more and better data-driven decision-making. Some incorporate subnational needs, others focus on population and family planning; few address both. Notable examples include:

• ¡ The World Bank and WHO, with input from several agencies and countries, have

developed a Global Civil Registration and Vital Statistics (CRVS) Scaling Up Investment Plan. It covers activities over a 10-year period from 2015 to 2024, with the goal of universal civil registration of births, deaths, marriages, and other vital

• events. The plan explicitly focuses on the provision of reliable data on population

size and distribution at all levels, trends in fertility, patterns and causes of mortality, emerging health threats, high-risk groups, and the ability to track health progress and the health status of population at national and subnational levels.24

• ¡ WHO’s Health Metrics Network and Health Data Collaborative provides a wealth
• f resources intended to support greater reliability and use of data in support of

country achievement of their own health and development plans and the SDGs.

• ¡ The United Nations Foundation has supported two initiatives—Data2X and Big

Data for Social Good. Data2X is a collaborative technical and advocacy platform dedicated to improving the quality, availability, and use of gender data to close gender data gaps, promote expanded and unbiased gender data collection, and use gender data to improve policies, strategies, and decision-making in support of gender equality.25 GSMA is an international mobile technology organization that spearheaded Big Data for Social Good, which utilizes anonymized mobile data from 100+ countries to monitor, assess and respond to humanitarian crises.26

• ¡ The Results for Data Initiative was launched by the Development Gateway to better

understand how local-level development actors actually collect, share, and use results data to inform government and agency leaders and development programs.33 After speaking with over 450 government officials, donor representatives, and implementer staff, they have begun to put their insights into action, with country- specific assessments in Ghana, Tanzania, and Sri Lanka.

• ¡ The Global Partnership for Sustainable Development Data is multi-stakeholder

network of more than 150 data champions harnessing the data revolution and

13 working to fill critical data gaps and ensure that data are accessible and usable in ending poverty.27

• ¡ The FP2020 Performance Monitoring and Evaluation Working Group28 was

prioritized within the FP2020 partnership from its inception. The group focuses on

• ngoing national and global monitoring of progress against FP2020 goals and

country commitments. It also aims to promote data use, provide technical assistance to governments, and advance the measurement agenda.

• ¡ In September 2017, The Bill and Melinda Gates Foundation launched the

innovative and interactive Global Goalkeepers initiative pairing data with compelling graphics, case studies, testimonials, and interactive web content. The content rests on the 2017 report on SDG progress and tracks 18 data points, including family planning, that the foundation believes are “fundamental to people’s health and well-being”.29 From Surveillance to Precision, Understanding, and Use The literature review and interviews also focused on how SNGs apply national and subnational data to policy and program development. Barriers to effective use of data exist at all stages in the cyclical nature of evidence-based decision-making at subnational levels (Figure 1). Key informants consistently underscored the need to improve the use of data by SNGs at each phase to improve the health and well-being of their localities. Figure 1. Phases of Data Use for Decision-making

Adapted from the Results for Development Initiative, Retrieved on August 25, 2017 from http://www.developmentgateway.org/blog/using-pdia-put-data-action.

14 They also highlighted constraints SNGs experience as they collect, receive, interpret and use data to inform and guide responses to needs for contraceptive information, services, and supplies among their constituencies. A number of themes emerge and support the recommendations presented in the discussion section.

• Precision. There is consensus that the DHS are critical to setting national priorities,

surfacing inequities within countries, and enabling comparisons among countries. Equally important are HMIS and DHIS2, which are both widely used throughout sub- Saharan Africa. Yet, as decentralized governance has taken hold in countries in this region, the limits of DHS, HMIS, and DHIS2 data are clear. Efforts to adapt to the limitations of such systems were described. As one informant noted, “Now there is more thirst, more hunger, for data that is applicable to states and local governments, and they want to support the Bureau of Statistics to generate subnational data.” Suggestions for addressing this hunger take many forms. Responding to subnational needs can be as simple as data analysis to produce contraceptive use trends or as complex as satellite images illustrating population density in specific geographies. Regardless, there is consensus that in order for SNGs to manage health services effectively and efficiently and for them to address population issues through these means, they need better information on the specific needs facing specific sociodemographic groups. They recognize the potential of mobile data and other technology to accelerate data collection and analysis and enable them to respond more accurately and quickly. In terms of family planning, meeting these data needs would strengthen the rationale for task-sharing and tailored outreach efforts and improve the accuracy of localized costed implementation plans, contraceptive procurement, and budget requests. Ethiopia is one notable example

• f how this can be done at all levels of government—from national to subnational.30

For the most part, current data collection systems are far from precise or reliable at the subnational level. With notable exceptions—those countries using DHIS—many countries in sub-Saharan Africa rely on paper-dependent systems for tracking patient and service information. These systems are prone to human error, inadequate personnel, and loss to competing priorities. One sub-Saharan informant noted that even when a health

15 issue, such as post-partum family planning is a government priority and provisions are made to include it in data collection, there is no clear instruction to subnational staff and relevant data collection points (e.g., columns in patient files) remain blank.

• Understanding. Informants raised a number of issues related to the ability of SNGs to

adequately interpret data in ways that are adequate and applied. Few officials within SNGs are trained statisticians or have statisticians on their staff. Education levels and data literacy are often cited as deficiencies, yet equally important are language barriers (i.e. facility with written English, French, or Portuguese), competing demands on time for reading and analysis, and poor presentation of information. Smaller SNGs also rarely have the multiple data sources, capacity, and resources needed to apply national and regional data to SNG contexts and assess performance and impact over time or in contrast with other SNGs. Even when subnational data are available, SNGs have varying degrees

• f skill in interpreting or questioning them within their own context and the license to

question them based on their own experience. Moreover, there is a tendency to focus only

• n official government population and health data when in reality a country’s ministry of

health rarely is the sole provider of services.8 This is especially the case with family planning, where the private sector plays an important and, in some countries a major, role in the provision of services and supplies. Finally, whereas newer, faster ways of collecting and providing data fill some national government and SNG needs, little attention has been paid to assessing them against the well-known and well-trusted data sources such as the DHS.

• Utilization. Governments, civil society, and donors are increasingly paying attention to

the ability of SNGs to use data effectively. For example, in 2012, Palladium developed a comprehensive eight-step intervention that moves data producers and users from examining and improving the context in which data are collected to data use monitoring, evaluating and dissemination.31 It is important to recognize that SNG data use is frequently driven by instructions from the national government and is highly proscribed. SNG data use can also be highly dependent on which donors or international agencies are active within a subnational geography and which data they deem most applicable to their

16

• wn needs.8 When focusing on improved use of data SNGs should be recognized as

having three roles: contributors, consumers, and communicators. According to one sub- Saharan African informant, “data use [by SNGs] would improve dramatically if they were consulted from the beginning—from formulation of the questions and identification

• f the survey population through analysis and interpretation.” The ability to complete this

cycle, however is often held back by challenges to strengthening SNG public health performance overall: most notably lack of continuity of leadership or personnel.32 Officials may be reticent to use or share data if the findings reflect negatively on their

• performance. Finally, a focus on family planning, in particular, may impede use.

Especially in smaller SNG units there may be issues around privacy and the possibility of data being used to stigmatize certain health behaviors or outcomes. DISCUSSION The findings in this assessment of the availability, reliability, and utility of data for SNG decision-making surrounding population, health, and development issues suggest the need for retooling existing mechanisms to fit SNG realities. At the heart of this analysis is a desire to identify how SNGs can use data to deliver results against health and development plans and commitments, such as those FP2020 commitments made and reaffirmed at the 2017 London Summit and those reflected in the SDGs. If existing national data systems are effectively adapted to the needs of the subnational level and are tested through implementation, SNGs can rely on these data systems and the evidence they produce to respond to and integrate national policies at the subnational level, develop their own subnational policies, and assess the results of policy decisions and implementation over time. That said, as one informant noted, efforts to improve the situation at the SNG level must be realistic, cost-effective, and the benefits made clear: “How much do you want to invest to make the statistics system work? Will investments here mean that other improvements are not made or commodities are not purchased?” This preliminary analysis of SNG data and needs presents an opportunity to mobilize existing resources for data improvement and use in improving women’s access to and use

17

• f contraceptive information, services and supplies in sub-Saharan Africa. Seven key

recommendations were established for national and subnational data producers, including researchers, data managers, and donors, to address the complex needs of SNG data users. These recommendations are based on a synthesis of ideas gleaned from the literature, KIIs, and adaptation of the national recommendations of others to the SNG context.

• 1. ¡ Improve systems for refuting data or “truth-testing”. Beyond the existence and

application of subnational data, one of the core challenges for subnational researchers and decision-makers is the management of divergent data. Informants described scenarios where data from one subnational source would contradict data from another

• source. Development of improved systems, including data systems that can collate

and assess divergent data from extant data systems, as well as training for subnational data producers and users to understand the limitations of different data sources, will serve as important first steps toward improving SNG data and their use.

• 2. ¡ Overcome inconsistencies, ensure interoperability. The many data improvement

efforts noted above shed light on the need to address how best to standardize certain critical measures, including those that apply to family planning. This would aid SNGs in following indicators across various surveys and enable comparability within and between countries. Improved interoperability and the extent to which systems and devices can exchange data, and interpret shared data could also benefit national and SNG data use.

• 3. ¡ Present information strategically. As researchers who leverage data across sub-

Saharan Africa for various measurement and monitoring purposes, there is a shared responsibility to disseminate findings of these data in strategic ways. It is important to make data analysis worthwhile by providing recognition and professional incentives for SNGs to review and use data in ways that support national and global goals.33 Equally valuable is the communication of data by sharing findings through means that are actionable, easier to understand, and relevant to the demands placed on local

• policymakers. Strategic presentation also requires the tailoring of data to local needs
• r audiences. As one informant noted, “For a district political leader, the interest is

value for money and [if it] will help him politically. Does it resonate with his or her

18 constituency?” There is need to focus on the cyclical nature of data for decision- making by involving SNGs and their communities throughout the data collection, measurement, and analytic processes.

• 4. ¡ Put data in context. While this analysis focused primarily on population and health

data and government responsibility, it is essential to analyze findings in the context of

• ther factors that would impede or speed progress. These include financial and

private-sector information, operations research, anecdotal evidence, and in the case of family planning, systems research, commodity and training information and gender- and youth-specific data. Moreover, improvement efforts must recognize that SNGs have challenges specific to the scope of their mandate and authority. Their success is also highly susceptible to key challenges resulting from limited leadership continuity, absenteeism, recidivism, and low visibility and status of public health work.

• 5. ¡ Adopt new technologies for subnational collection and presentation. A common

denominator in improving the availability, reliability and use of data is the antiquated and cumbersome nature of hand-entered, paper registers to collect client and service

• statistics. While internet access in many remote areas within sub-Saharan Africa is

poor, cellular and satellite technology is ubiquitous and could be leveraged to improve data collection, synthesis and dissemination. Communication techniques, such as those used in the Global Goalkeepers29 could be adapted to the technology available and help meet the need to incentivize SNGs in their collection and use of

• data. Online platforms are increasingly used at the national level to connect public

servants at all levels34 and could be adapted to include SNG data and progress. For example, in Uganda, Samasha Medical Foundation developed the Motion Tracker, an innovative approach to galvanizing civil society support for and involvement in achieving FP2020.35 The system is a customized, dynamic framework for strengthening accountability and driving action by keeping commitments visible and highlighting progress and engaging government and civil society. The tracker includes an online, user-friendly tool for visualizing action and progress towards family planning commitments, though this approach can be translated to other areas.35

• 6. ¡ Pair investments in improving SNG data use with those improving data
• presentation. While training on basic data literacy is needed, it is only one aspect of

19 skills-building. It is important to match initiatives in this area to the realities of SNG data utility. One informant stressed this point, “We need to put ourselves in the place

• f a decision-maker. If you are undertaking a survey or providing data, you need to

consider what you would and could do if you were the governor of Homa Bay [in Kenya] and you have a high HIV prevalence rate? What more information would you need to take action? In other words, don’t present information that is challenging to understand and that doesn’t relate to what is actionable.” Assessments of audience needs and interests are critical to deciding which data to present and how best to present them. This type of assessment has been used successfully throughout Advance Family Planning (AFP) focus countries to support SNGs policy and program development and implementation.4

• 7. ¡ Recognize that family planning data are political. No data are neutral, but

information related to reproductive health can be especially controversial if they are found to be unreliable or they do not fit with a decision-maker’s preconceptions or

• values. Moreover, they often put a decision-maker at risk of inciting negative

reactions from their communities and, in some cases, their own government. Efforts to improve data availability, reliability and use by SNGs will need to place equal weight on helping local officials represent sensitive issues appropriately and with

• confidence. Further, when new data systems are introduced or significant

methodological changes are made to existing data systems, decision-makers and local leaders who are likely to use the data should be informed of these changes. Early engagement of SNGs in these changes will facilitate data trust and translation to evidence-based policy and practice. ¡ These recommendations are rooted in the diverse experiences of informants and are intended to guide action for the improvement of data systems that can comprehensively address the needs of subnational governments to measure, evaluate, and monitor the population, health, and development issues of their geographies. This research is limited in that informants were purposively selected based on their expertise and knowledge in the area of family planning data and policy in Africa. Further research is needed to evaluate individual country experiences with data for subnational decision-making

20 related to population health issues and to explore the implementation of these recommendations in sub-Saharan African contexts. ACKNOWLEDGEMENTS The authors acknowledge the contributions of the 25 global health researchers and practitioners who provided invaluable insight through key informant interviews, research assistance of Basile Moreau, and the support of the Bill and Melinda Gates Foundation, the William and Flora Hewlett Foundation and the David and Lucile Packard Foundation. We also wish to recognize the many government officials throughout sub-Saharan Africa and the world who dutifully record information, care that it is accurate, and use it to improve the lives of people in their communities.

21 REFERENCES ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡

1 Manor J. (1999). The Political Economy of Democratic Decentralization. World Bank. 2 Parker A. (1995). Decentralization: The Way Forward for Rural Development? Policy

Research Working Paper, 1475. World Bank, Washington, D.C.

3 Bossert, T. (1998). Analyzing the decentralization of health systems in developing

countries: decision space, innovation and performance. Social Science and Medicine, 47(10), 1513-1527.

4 Gillespie D, Hawks Cuellar M, Whitmarsh S, Bodenheimer A, Karklins S. (2016).

Connecting Global Goals to Local Priorities. Global Health Governance, 10(3), 54-65.

5 Global Financing Facility. (2017). About us. Retrieved August 25, 2017 from

https://www.globalfinancingfacility.org/introduction.

6 GFF. (2016). Resource Mobilization Annual Update, Global Financing Facility Fourth

Investors Group Meeting, 3-4 November, 2016. Retrieved September 22, 2017 from www.globalfinancingfacility.org.

7 The World Bank. (1998). Rethinking Decentralization in Developing Countries. 8 Rohrer, K. (2016) Strategizing for health at sub-national level. World Health

Organization.

9 Chase L. (2017). Our Saint of Rational Self Interest. ELLE, 241-242. 10 Sustainable Development Solutions Network. (2015). Data for Development: A Needs

Assessment for SDG Monitoring and Statistical Capacity Development.

11 Institute of Medicine. (2011). Committee on a National Surveillance System for

Cardiovascular and Select Chronic Diseases. A Nationwide Framework for Surveillance

• f Cardiovascular and Chronic Lung Diseases. Washington (DC): National Academies

Press; Existing Surveillance Data Sources and Systems.

12 Dunning C, Rose S, McGillem K. (2017). Implementing Ownership at USAID and

MCC: A US Agency-Level Perspective. Center for Global Development.

13 United Nations. (2015). Transforming our world: The 2030 agenda for sustainable

development.

14 Fryatt RJ, Bhuwanee K. (2017). Financing health systems to achieve the health

Sustainable Development Goals. Lancet Global Health. http://dx.doi.org/10.1016/S2214- 109X(17)30263-2.

15 Ahmad J, Devarajan S, Khemani S, Shah S. (2005). Decentralization and Service

• Delivery. World Bank Policy Research Working Paper 3603. World Bank.

16 ¡Rondinelli DA, Nellis JR, Cheema GS. (1983). Decentralization in developing

countries: A review of recent experience. World Bank Staff Working Paper 581. Management and Development Subseries 8. World Bank. ¡

17 Schorr L. (2016). Reconsidering Evidence: What it Means and How We Use It.

Stanford Social Review.

18 Garrido M, Fellows M. (2017). Access to Information and the Sustainable

Development Goals, Development and Access to Information. International Federation of Library Association and Institutions and The Technology & Social Change Group.

19 World Health Organization. (2008). Health Information Systems. 20 Pullum TW. (2014). The Demographic and Health Surveys: A Program and an

• Institution. USAID. Presentation on November 12, 2014 to students and faculty in the

22 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ Department of Population, Family and Reproductive Health at the Johns Hopkins Bloomberg School of Public Health.

21 Anderson JE, Cleland JG. (1984). The World Fertility Survey and Contraceptive

Prevalence Survey: A Comparison of Substantive Results. Studies in Family Planning, 15(1), 1–13.

22 MEASURE DHS. (2012). Demographic and Health Surveys. Calverton: ICF

• International. Retrieved September 17, 2017 from: http://www.measuredhs.com/.

23 World Health Organization. (2014). Health Facility and Community Data Toolkit. 24 World Bank Group. (2014). Global Civil Registration and Vital Statistics, Scaling up

Investment Plan 2015-2024.

25 Buvinic M, Furst-Nichols R, Koolwal G. (2014). Mapping Gender Data Gaps:

Executive Summary. Data2X.

26 GSMA. (2017). GSMA Launches Big Data for Social Good Initiative. Press Release.

Retrieved June 5, 2017 from https://www.gsma.com/newsroom/press-release/gsma- launches-big-data-for-social-good/

27 Partnerships for SDGs. (2017). Annual Report. Retrieved September 28, 2017 from

https://sustainabledevelopment.un.org/partnership/?progress&id=600.

28 FP2020. (2017). Data Measurement Hub; The Performance Monitoring and Evidence

Working Group. Retrieved September 22, 2017 from http://www.familyplanning2020.org/microsite/measurement-hub/pme-wg.

29 Bill and Melinda Gates Foundation. (2017). The Stories Behind the Data. Retrieved

September 22, 2017 from http://www.globalgoals.org/goalkeepers/.

30 USAID. (2012). Health care financing reform in Ethiopia: improving quality and

• equity. Washington, DC.

31 Nutley T. (2012). Improving Data Use in Decision Making: An Intervention to

Strengthen Health Systems. Measure Evaluation. Special Report.

32 WHO. (2017). Public health performance strengthening at districts:

Rationale and blueprint for action. Proceedings of a Bellagio Conference November 21- 24, 2016. Retrieved August 17, 2017 from www.who.int/alliance- hpsr/bellagiowhitepaper.pdf.

33 Homer D. (2016). Doing Data Differently: Lessons from the Results Data Initiative.

From Poverty to Power. Retrieved June 9, 2017 from https://oxfamblogs.org/fp2p/doing- data-differently-lessons-from-the-results-data-initiative/.

34 Scott R, Witter L. (2017). Bridging Governments’ Borders. Stanford Social Review. 35 HP+. (2017). The Motion Tracker: Monitoring Local Action to Hold Governments

Accountable for Global Commitments. Retrieved September 9, 2017 from http://www.healthpolicyplus.com/ns/pubs/2099-3169_TheMotionTracker.pdf. ¡