Data revolution for SDGs

Last month, I attended two workshops organized by the Pakistan Bureau of Statistics (PBS) on user-producer dialogue to sense data needs and calibrate with statistical products – a commendable step towards making official statistics demand-driven.

This reminded me of the UN initiative on data revolution to fully integrate statistics into decision-making and promote open access and use of data. Data matters because it can see, hear, feel, whisper, and even shout; it reflects human life with its joy, sorrow, and success; it tells a lot if we can give it an honest voice.

Planners face many challenges and contradictions, and the future will not be less challenging, but it will remain promising, as data accessibility has no limits. Good data strengthens the strategic foundations of planning and gives confidence to turbocharge operations, while a lack of data clogs the processes and undermines trust in statistical systems. The dual question at hand is about data ecosystem and accessibility that allow users to access, understand, analyze, use, reuse, and link data to identify current and future challenges.

In a global context, the endorsement of the amendments to Pakistan’s statistics law in 2011 and the development of the national strategy for the development of statistics (NSDS) provided a very modernized legal framework and strategic direction. Yet implementing the law and the NSDS remains a challenge due to fragmented efforts and bureaucratic processes. The current national statistical system (NSS) follows the textbook traditional supply-driven approach. It is centered at the PBS as a strong and capable statistical office and in provincial statistical bureaus but there is a dire need for modernization and capacity. In the data landscape, the private sector is left behind, although it is a major partner as a provider and user of data. According to the World Bank’s Statistical Performance Indicator (SPI) measuring statistical services, products, sources and infrastructure, Pakistan scored a 61 per cent ranking globally after India (67), Malaysia (65), Egypt (57), and Turkey (29).

On the supply side, although various statistical reports are published, access to digital data and timeliness of dissemination remain the key challenges. In addition, updated, disaggregated and user-friendly is not easily available; for example, the latest available data on maternal mortality is from 2019, contraceptive prevalence is from 2018 and unemployment is from 2021.

Data at the district level is rarely available which is particularly significant considering the wide disparities across districts. For example, Hyderabad district is performing four times better than Rajanpur district on the UNFPA Reproductive Provisional Health Index (RHI). More broadly, Pakistan produces data on few SDG indicators; according to the SDGs unit, only 60 indicators are covered by official statistics sources.

The completion of a digital census in Pakistan was a great step forward from a supply perspective, but a full digital transformation including digital dissemination must be introduced to eliminate slow and hard-wired procedures. In a high-speed world, many users are not interested in reading long statistical reports and detailed tables and are instead more attracted to synthesized and customized information.

Pakistan’s statistical sources are not on the primary list in search engine results even for simple searches on reproductive health and mortality indicators. This means dissemination and presentation functions require modernization and optimization. On the research front, the data ecosystem is a key enabler for R&D. Pakistan spends only 0.16 per cent of its GDP on R&D, while India spends four times more and Turkey spends eight times.

On the demand side, there is a huge user demand for digital data, but there may be a lack of capacity to do their own analysis. The future demand for data goes beyond the capability of the current statistical system and data ecosystem; it speaks to artificial intelligence and machine learning. Such a technology-driven future will demand better functional literacy that equips people with the capacity for effective functions for self and community development. Also, having statistical analysis on the education of girls and boys can be an important step so their skills fit the future labor market needs. For example, data can tell us why women are stuck in the use of traditional contraceptives and not transiting to modern methods.

At the political level, creating a data-driven culture requires that the national statistical system is trusted as professionally independent, technically capable, politically unbiased, and away from any form of interference. Trust is earned not imposed; good practices reveal that statistical authority should be independent. Statistical independence was defined by the UN Fundamental Principles for Official Statistics as having full control over methods and technical decisions including what, when, and which level of details of the disseminated statistics.

No other authority should ‘approve or disapprove’ a statistical product, given the administrative, organizational, and accountability arrangements of the statistical office within the government structure. Statistics are ‘public goods’ to which everyone including institutions, politicians, government, civil society and the public at large has access at the same time and the same level of detail. Statistics are a pointer for public debate, not the opposite.

Pakistan is in dire need of a state-of-the-art, proactive, digitalized, and innovative data ecosystem to provide a digestible presentation of relevant and real-time information. It will create a culture of public debate on data, empower evidence-based decision-making, provide an analytical and research environment, contribute to effective monitoring of SDG indicators, and enforce the 2017 Act on the right of access to information. It is time to modernize.

The writer is the UNFPA representative in Pakistan.