5 CHAPTER 1. INTRODUCTION 1.1 WHY ARE DATA AND DATA ECOSYSTEMS IMPORTANT? U seful, accurate data is essential for policymaking that will support Myanmar’s socioeconomic development and strengthen public service provision. In making decisions, policymakers must identify, understand, and prioritize policy issues, assess and select potential solutions, and analyze the impact of interventions. At each of these stages, data can provide crucial evidence that helps make sense of complex systems and ensures value for money, responsiveness, and equity in decision-making. In recent years, the importance of data has increasingly been recognized as part of the movement towards evidence-based policymaking. At its heart, evidencebased policymaking emphasizes that “what matters is what works,”1 with the application of rigorous data analysis to understand issues and the impact of policies. The movement has been reflected in the policies of the government of Myanmar with the recognition that policymaking should be “rooted in strong evidence.”2 Today, data is ubiquitous, with technology providing new opportunities for its collection, storage, use, and analysis. Of course, data collection in Myanmar is nothing new: for centuries, administrative ledgers, geographic surveys, and censuses have provided muchneeded tools for government to function. But the future is one of more data, gathered in innovative ways, such as from internet-connected sensors or public input via smartphones. Modern infrastructure and consumer devices provide the capability to automatically collect huge volumes of data, and cloud computing and bigdata analytics provide new opportunities for the storage and analysis of huge datasets. The greater availability of data is not an end in itself, however, and it will not strengthen decision-making if it overwhelms policymakers, if it is untrusted or of poor quality, if it is not available in formats that people can easily use, or if policymakers do not have the skills or tools to make good use of it. Strengthening decision-making requires that we consider the data ecosystem as a whole (figure 1.1), rather than focusing myopically on “lots of good data.” Fundamentally, does the right data, in the right format, get to the right people with the right skills and at the right time to support effective decision-making? Consideration needs to be given to the full lifecycle of a datapoint, including: z Collection. Why, what, how, and when is data collected, and who collects it? How is the quality of data assured? z Storage. What data is stored, how, and by whom? z Analysis and use. What data analysis is carried out and by whom? Who uses data and how? z Governance. What is the framework that governs data standards, who has access to data, and how is data secured? How do those staff working within the local data ecosystem perceive data, and what skills, training, and experience with data do they have? If policymakers are to base decisions on data, that data must be of adequate quality to support quality decisionmaking. Conversely, if decisions are based on poorquality data, there are significant risks that policy will not meet its aims. In examining the data ecosystem, special attention should be given to the quality of the data, so that there is confidence in its accuracy and, thus, its usefulness. Is the data timely, complete, accurate, consistent, and understandable (figure 1.2)? The data quality of official statistics in Myanmar has ranked at the bottom of ASEAN countries3 and second lowest in a recent World Bank analysis of ASEAN statistical capacity.4 Beyond the quality of the data, it is necessary to consider the efficiency and effectiveness of the data ecosystem. For example, data collection can be expensive, and care should be taken to collect the right data, and in a way that is the least resource intensive while maintaining data quality. In many sectors in Myanmar, data collection and storage are still paper based, with some forms of data dating to the British colonial and Burma Socialist Program Party administrations. As a consequence, some data may serve little purpose in informing decisionmaking, and paper-based data impedes more complex computer-facilitated analysis.