Imagine embarking on a journey to an unknown destination only to find halfway through that two-thirds of your map is missing – this is the reality for gender statistics on the heels of the SDG Summit at the United Nations which marks the midpoint of the 2030 Agenda. In a recent working paper, “Missing SDG Gender Indicators”, we examine coverage of gender data in the SDG monitoring framework and find that only a third of gender related SDG data are available. Moreover, we try to unpack what drives this lack of data.
Of the 231 indicators in the SDG monitoring framework, 50 indicators are gender-related: 14 are Goal 5 (Gender equality) indicators, and another 36 are either gender-specific indicators in other goals or indicators requiring sex-disaggregation. For each of these 50 gender-related indicators, we examine whether a country has at least one data point between 2016 and 2020, and find that on average, only a third of the countries do. In fact, over 90 percent of the world’s population lives in a country where more than half of the 50 indicators were not available for any year in this 5-year period.
In a regional comparison (see the figure below), South Asia stood out with the highest SDG gender indicator coverage at 36 percent availability – albeit only 5 percentage points higher than the global average.
To be sure, this noted lack of gender data is well-documented. In our paper, we seek to unpack the reasons why gender data may not be available.
Poor data processing and reporting is a cause of missing gender data
Inadequate data collection is an undeniable reason for missing gender data. However, we find that even when data are collected, they are often not processed for sex-disaggregation.
For example, for the SDG indicator which measures the proportion of population covered by social protection floors/systems (SDG indicator 1.3.1), around 80 percent of countries have a recent value for the population, but only 8 percent of countries have a sex disaggregated data point. A less drastic example is related to early childhood education (SDG indicator 4.1.1): 64 percent of countries have a population estimate for this indicator but only 53 percent have an estimate disaggregated by sex.
Of the 50 indicators, 32 indicators are sex disaggregated. On average, 43 percent of these indicators have population estimates, but only 31 percent have corresponding sex-disaggregated data. If countries that had a population estimate also had a sex-disaggregated estimate (which is almost always feasible from what we know of these data sources), the Sustainable Development Goal gender coverage rate would increase to 43 percent.
This aligns closely with one of the main messages of the World Development Report 2021: Data for Better Lives that gathering more data is not the only answer, and realizing greater value from data by extracting more from what is already collected is essential to drive policy decisions.
Strong statistical systems, not high-income level, determines gender data coverage
Contrary to expectations, in a comparison of SDG gender indicators coverage by income level, high-income countries do not have higher coverage. In fact, GDP per capita is neither associated with better coverage of gender statistics nor overall statistics in the UN SDG database (see the figure below).
Yet, SDG gender indicator coverage is very strongly correlated with Statistical Performance Indicators and Index (SPI) – the World Bank’s new official tool to measure country statistical capacity (Dang et al. 2023) (see the figure below). This tells us that better statistical systems are a major part of the solution.
Country-level advocacy can catalyze better coverage of gender data
Lastly, many countries over- (and under-) perform in reporting gender SDGs compared to what would be expected conditional on statistical strength, as shown in the figure below. For example, Serbia reported on 58% of the SDG gender indicators. Based on the SPI overall score of the country, it was expected to produce only 40% of the SDG gender indicators. This 18 percentage points difference translates to 9 additional indicators reported compared to expected coverage.
These differences in coverage suggest that country-level advocacy and focus can yield wins in Sustainable Development Goal gender indicator coverage, conditional on statistical capacity.
The concern of missing gender data can be addressed in part by targeting low-hanging fruits like processing existing data for complete sex-disaggregated information. And to further improve availability of gender data, country level advocacy and improvements in statistical systems will be imperative.