Gender Data Gaps: A Comparison of Three Regions

by Tawheeda Wahabzada and Eric Swanson
Open Data Watch

In April 2021, Data2X and Open Data Watch, in collaboration with the UN Economic and Social Commission for Asia and the Pacific (ESCAP) launched Bridging the Gap: Mapping Gender Data Availability in Asia and the Pacific. This was the third regional review in the Bridging the Gap series. We hope that the results of these project will be used and reused to monitor progress towards the Sustainable Development Goals (SDGs) and to identify indicators needed to monitor national development strategies with a gender lens.

The Bridging the Gap studies have examined gender indicators available from international databases and national databases in 25 countries across three regions: 15 countries in Sub-Saharan Africa, 5 in Latin America and the Caribbean, and 5 in Asia and the Pacific. International databases included the websites of SDG custodian agencies and the SDG Global Database maintained by the United Nations Statistics Division. The principal national databases were found on the websites of national statistical offices and relevant ministries within the national statistical system.

In both international and national databases, the following types of gender data gaps were tabulated:

Availability of disaggregated data: Are data available in the last 10 years and are the indicators disaggregated by sex and other characteristics?
Adherence to standards: Do the available data conform to international standards, particularly those of the UN Statistics Division’s SDG metadata repository?
Timeliness and frequency: How timely are the data, and how many annual observations are available?

For each regional Bridging the Gap a list of gender indicators was compiled from SDG indicators classified at the time of the study as Tier I (well-defined methodology and widely available) or Tier II (well-defined methodology but not widely available) and from other indicators recommended by UN Women.[1] The study of Sub-Saharan Africa included 104 gender-relevant indicators, 68 of which are SDG indicators; the study of Latin America and the Caribbean included 93 gender-relevant indicators, 84 of which are SDG indicators; and the study of Asia and the Pacific included 98 gender-relevant indicators, and 93 of which are SDG indicators. To make comparisons across the three regions, this report utilizes the core set of 68 SDG indicators that appeared in all three studies. This report only focuses on findings from national databases.

Data availability and disaggregation

Throughout the Bridging the Gap regional studies we found a shortage of gender data that limits our knowledge of the status of and well-being of women and girls in countries throughout the world. In national databases, 50 percent of the indicators had sex-disaggregated data, 18 percent lacked sex disaggregation, and 32 percent lacked any observations in the preceding 10 years.

Figure 1: Average indicator availability in national databases (68 indicators)

Sub-Saharan Africa

The Bridging the Gap study included 15 countries from Sub-Saharan Africa. Assessments were carried out between June 2018 and September 2018. On average Sub-Saharan Africa has the smallest proportion of indicators with sex-disaggregated data: 45.6 percent of the indicators in national databases, and the largest proportion, 40.3 percent, lacking any data. Malawi had the highest proportion of sex-disaggregated data, although Nigeria had the fewest indicators lacking any data. Almost 60 percent of the indicators in Lesotho’s national databases lacked any data.

Figure 2: Availability of 68 SDG Gender Indicators in 15 Sub-Saharan Africa Countries

Latin America and the Caribbean

Assessments were carried out in five countries from Latin America and the Caribbean between June 2019 and August 2019. On average, 54.7 percent of the SDG gender indicators in national databases of the five countries included sex-disaggregated data, while 24.1 percent lacked any data at all. Costa Rica had the largest proportion with sex-disaggregated data and the smallest proportion with no data. Jamaica with the smallest proportion of sex-disaggregated indicators also had the largest proportion of indicators with data lacking sex-disaggregation.

Figure 3: Availability of 68 SDG Gender Indicators in 5 Latin America and Caribbean Countries

Asia and the Pacific

Assessments in Asia and the Pacific between May 2020 and August 2020. The five countries from Asia and the Pacific had the highest proportion of sex-disaggregated indicators in their national databases. On average 60.3 percent of the SDG indicators were available with sex-disaggregated data; 24.7 percent had data but lacked sex-disaggregation; and 15.0 percent had no data. The Philippines had the highest proportion of sex-disaggregated indicators and the smallest proportion lacking any data. Samoa had the largest proportion of indicators lacking data.

Figure 4: Availability of 68 SDG Gender Indicators in 5 Asia and Pacific Countries

Data availability by domain

Indicators included in the Bridging the Gap studies were grouped into six topical domains. Among the 68 SDG indicators, the largest domain is health, with 22 indicators; the smallest is public participation with 4. In all three regions, the environment domain has the largest gap in sex-disaggregated data, where 6.7 percent of indicators are available with sex-disaggregated data in Sub-Saharan Africa, 8.6 percent in Latin America and the Caribbean, and 5.6 percent in Asia and the Pacific.

Figure 5: Sex-disaggregated indicators in national databases by domain (%)

In Sub-Saharan Africa, the largest proportion of sex-disaggregated data is in the public participation domain, where 63.3 percent of indicators are available with sex disaggregation. This is followed by the health domain, where 54.5 percent of indicators are available with sex-disaggregation. In Latin America and the Caribbean, the largest proportion of sex-disaggregated data is in the economic opportunities domain, where 67.5 percent of data are available with sex disaggregation. This is followed by public participation, where 65 percent of data are available with sex disaggregation. Finally, in Asia and the Pacific, the largest proportion of sex-disaggregated data are in the public participation domain, where 80 percent of indicators are available, followed by the economic opportunities and human security domains, where 68.3 percent of indicators are available with sex disaggregation.

Adherence to standards

Adherence to international standards refers to whether the available SDG indicators conform to the standards and definitions listed in the SDG metadata repository. Indicators that do not meet standards of the metadata repository for construction or presentation were classified as non-conforming indicators. Examples of non-conforming indicators in the three Bridging the Gap regions include the following:

Sub-Saharan Africa: SDG 3.3.3 Malaria incidence per 1,000 population: In Botswana’s national database, data are on reasons for out-patient and in-patient attendance, which includes malaria as a line-item.
Latin America and the Caribbean: SDG 11.2.1 Proportion of population that has convenient access to public transport: In Jamaica’s national database, data are on proportion of people that use public transport, not people with convenient access.
Asia and the Pacific: SDG indicator 3.9.3 Mortality rate attributed to unintentional poisoning: In Samoa’s national database, aggregated data are on deaths by “injury, wounds, poisoning, & certain other consequences of external causes.” Data on poisoning is grouped with other causes of death; furthermore, it is uncertain whether the act is unintentional.

Across the six topical domains, public participation has the largest share of conforming indicators with sex disaggregation (56 percent), and the environment has the largest share of conforming indicators without sex disaggregation (17.7 percent). The education domain has the largest share of non-conforming indicators with sex disaggregation (27.4 percent) and the environment domain has the largest share of non-conforming indicators without sex disaggregation (30.3 percent).

Figure 6: Conforming and non-conforming indicators by domain

Looking at the availability of conforming and non-conforming indicators by region, we see that in Sub-Saharan Africa, 28.7 percent of the indicators are conforming with sex disaggregation; 16.9 percent are non-conforming with sex disaggregation, and 40.3 percent lack data. In Latin America and the Caribbean, 40.6 percent of the indicators are conforming with sex disaggregation; 14.1 percent are non-conforming with sex disaggregation; and 24.1 percent lack data. And in Asia and the Pacific, 51.2 percent of the indicators are conforming with sex disaggregation; 9.1 percent are non-conforming with sex-disaggregation; and 15 percent lacks data.

Figure 7: Conforming and non-conforming indicators by region

Timeliness and frequency

Timeliness

Timeliness of data refers to the time taken to go from data collection to publication. Timely data are needed to implement policies and monitor progress to the SDGs. Because of the time needed to plan and conduct surveys, there is typically a lag of two years or more between the nominal date of an observation and its availability in a national or international database. Indicators derived from digitized administrative records can be made available more promptly, if given priority.

The charts below show the distribution of the most recent observation available in the three regions. The countries of Sub-Saharan Africa and Latin America and the Caribbean show peak values in 2016, which reflects an effort at the end of the Millennium Development Goals (MDGs) period to document progress on important development indicators. The five countries of Asia and the Pacific have more recent data, with 37.4 percent of available indicators having an observation in 2018 and another 12.5 with observations in 2019.

Figure 8: Year of most recent observation by region

Because the assessments were carried out in different years, the lag between the last year of observation needs to be adjusted by the year of the assessment. The median year of observation for the fifteen countries of Sub-Saharan Africa was 2015, a difference of 3 years from the assessment in 2018. In Latin America and the Caribbean, the median year of observation was 2016 and the difference from the assessment year of 2019 was 3 years. And in Asia and Pacific with a median year of observation of 2017, the difference was again 3 years. However, the Africa region has a much thicker tail of many observations that were 6, 7, and 8 years old. Indicators from the Asia and Pacific region are heavily clustered in the two most recent years, but even here more than 80 percent of available indicators were 2 years old or older. In a world of rapidly changing events requiring up-to-date evidence for informed decision making, delays in publication reduce the value and usefulness of the data and old data are soon forgotten.

Frequency

Timeliness and frequency of data are closely related. Frequently measured data should be more timely. But even a timely observation may be of little interest if there is no way to measure change over time. Frequent observations, measured at regular intervals, are needed to monitor progress to monitor national development plans, the implementation of strategies and policies, and progress to the SDGs.

Here we look at data density as a proxy measure for data frequency: the more observations in a given time period, the more frequently they are measured and reported. As shown in the accompanying chart, many indicators are sparsely measured. There is no region where half of the available indicators have more than three observations in their national or international databases. The largest share of indicators in all three regions had between 1 and 3 observations over the 10-year period. Around 21 percent of the indicators in Asia and the Pacific and Latin and America and the Caribbean but fewer than 5 percent in Sub-Saharan Africa had more than 3 observations. Measured at regular intervals and using appropriate methods of interpolation or extrapolation, most of the SDG indicators should be available annually or more frequently.

Figure 9: Indicator density in national databases

Conclusion

Across the three regions, data gaps persist in the availability and disaggregation of gender data, adherence to standards, and the timeliness and frequency of data. With less than a decade left to fulfill the Sustainable Development Goals, timely and disaggregated data, measured at regular intervals, are needed to design and implement programs to address critical development issues and to monitor their progress. National policymakers, international agencies and bilateral donors rely on these data to identify needs and set priorities. A lack of gender data means that the needs of women cannot be adequately addressed.

Although these studies focused on the gender data needed to monitor the SDGS, data are also needed by national and local policymakers to formulate evidence-based, gender development policies, plans, or strategies. The Bridging the Gap studies have looked at national gender development plans and strategies and found that many lacked sufficient data to define their outcomes and monitor progress. Without data there is no accountability for results. Every development plan should, therefore, include a corresponding statistical development plan to ensure that critical gender indicators are available at every stage of policy creation and implementation.

Footnotes

[1] The Inter-Agency and Expert Group on the SDGs (IAEG-SDGs) has continually reviewed SDG indicators and their methodology. Over the course of the Bridging the Gap studies, IAEG-SDGs updated some gender indicators from Tier III (indicators that lack an established methodology) to Tier II or Tier I. Only Tier I and II indicators were included in the Bridging the Gap analysis. UN Women identified additional indicators, both from the SDGs and supplemental ones that, if sex-disaggregated, would provide additional insight. Indicators analyzed here are drawn from both sources.