Is the Pandemic Decreasing State Capacity?

While some countries overcame obstacles caused by the pandemic, others were pushed further behind in their efforts to collect critical socioeconomic data. Visiting Scholar Michael Harsch and Harvard student Alexandra Norris examine the latest trends in a follow-up to Harsch’s “Measuring State Fragility” post.

Niger farmers with cattle

By Michael Harsch and Alexandra Norris

As the world grapples with the economic and health consequences of COVID-19, another worrying long-term effect of the pandemic is becoming clear: growing state fragility. While governments are justifiably consumed with crisis management and disaster relief, the pandemic is jeopardizing countries’ statistical capacity by delaying the collection of fundamental demographic and economic information. If these delays become permanent, they could leave governments “partially blind” to the conditions in their countries and have severe negative implications for public service provision and long-term food security.

We use here a simple, innovative measure of informational state capacity to illustrate this largely overlooked impact of the global pandemic. The pandemic has disrupted the collection of crucial statistical data across the world; in particular, many agricultural censuses were postponed. Countries that already suffered from low statistical capacity proved to be vulnerable and the subset of states lacking up-to-date agriculture censuses, population censuses, and complete vital registries grew from twelve to fourteen countries. While this increase could ultimately be temporary, it illustrates the enormous challenges for maintaining stability in a post-pandemic world.

Measuring State Fragility

Fragile states are countries whose central governments lack the capacity to administer their territories effectively. Even before the pandemic, these countries struggled with political violence and providing basic socioeconomic opportunities to their populations. The pandemic has exacerbated these challenges.

There is a growing international consensus about the importance of informational state capacity for development outcomes and its usefulness as a proxy for overall state capacity. The UN’s Sustainable Development Goals (SDGs) for the year 2030, for instance, urge the international community to “develop measurements of progress on sustainable development that complement gross domestic product, and support statistical capacity-building in developing countries.”

While existing rankings of state fragility suffer from incomplete or poor-quality data about these countries, in the context of a research project on “islands of stability in fragile states,” we find that the best indicators of state fragility might be missing indicators. To assess and compare informational state capacity across the developing world, we leverage the latest World Bank data on the statistical capacity of low- and middle-income countries (excluding small states with fewer than 250,000 inhabitants).

Specifically, our measure assesses whether countries lack one or more of three crucial types of records: 1) a vital (birth and death) registry accurate within the past fifteen years, 2) a population census, and 3) an agricultural census (land use, production, etc.). Censuses are valid for one decade after completion. For this update, we corrected a small number of coding errors in the World Bank data (see documentation).

We use this information to classify countries as follows: states lacking up-to-date records in all three categories have very low statistical capacity; those missing two out of three types of records possess low capacity. These two groups of countries also qualify as fragile. By contrast, states lacking only one or none of these records have consolidated (two types of records) or high (complete records) statistical capacity; as such, they are not considered fragile. This measure of informational state capacity complements and improves upon existing metrics of state fragility: it is simple, transparent, and universally applicable.

Using more evocative terms, we can refer to states with very low statistical capacity as “collapsed” and to those with low capacity as “weak.” These terms are more likely to raise international awareness and will make it easier to mobilize resources to build capacity, though governments of such states may be reluctant to use them.

The animated map provides an overview of the global evolution of statistical capacity over the past fifteen years, the time period consistently covered by the World Bank data. (The World Bank started collecting this data in 2004 and later expanded it to the 1980s, but there are many missing values in the years prior to 2005.) From 2005 to 2019, statistical records generally increased, specifically in South and Central America, Southeast Asia, and parts of sub-Saharan Africa; exceptions were northern Africa, the Middle East, and the post-Soviet region, where the statistical capacity decreased during this period.

Yellow-Consolidated States; Orange-Weak States; Red-Collapsed States

[ View fullscreen interactive map ]

What Happened in 2020?

In 2020, the pandemic had a negative impact on many countries’ efforts to gather data. Concerningly, the number of countries with very low capacity, or collapsed states, increased from twelve to fourteen countries; this group now includes Burundi and Liberia. But even within this small group of countries, there is significant variation. Afghanistan, the Democratic Republic of the Congo, and Somalia either never collected any of the three records or all three have been outdated since at least 2005, highlighting the severity of the challenges faced by these countries’ governments. By contrast, Burundi and Liberia rejoined this group in 2020 after previously improving their statistical capacity, which might make it more appropriate to characterize them as being at risk of collapse.

Many countries were forced to delay their agricultural censuses—in some cases indefinitely. For example, Papua New Guinea, whose last agricultural census is from the early 1960s, planned to conduct a pilot census in the first half of 2020; yet due to travel restrictions related to the pandemic and a lack of funding, it suspended its plans. In Guinea in West Africa, preparations for its long-delayed agricultural census were scheduled to happen in March 2020 but were postponed. In a recent survey conducted by the UN, over two-thirds of countries whose agricultural censuses became outdated in 2020 reported that the pandemic disrupted their planning, data collection, or enumeration efforts.

This is a concerning development. Agricultural censuses are critical for understanding farming activity and capacity within countries, particularly in developing nations where agriculture can account for over 25 percent of GDP and the pandemic has already increased food insecurity. Censuses enable nations to take stock of their agricultural sectors, including land use and labor, which are crucial for a country’s food production. Even temporary delays can lead to permanent funding losses, particularly in a context of declining national budgets due to the global economic downturn. And growing distance from the previous census makes the analysis of trends more difficult and can complicate future censuses, as government experts move on to other jobs and institutional knowledge is lost.

Potato farmer kneeling in the field in Papua New Guinea

Yet these delays do not have to be permanent. For example, according to the World Bank data, we do not observe a decline in statistical capacity in West Africa during the 2014–2016 Ebola outbreak. This was partly due to timing—Guinea completed its 2014 population census in early April, days after the WHO confirmed the outbreak—but was also a result of strong international support, which enabled Sierra Leone to conduct a census in late 2015 and even include questions about Ebola.

In the current pandemic, success stories already exist. Gabon, Kenya, and Tanzania’s agricultural data collection efforts were interrupted by COVID-19 in 2020, but through the use of digital technology and additional training of enumerators, these countries were able to resume and complete data collection. These hopeful examples suggest that censuses in the developing world can be conducted safely during a pandemic.

Time for Global Action

As the UN is gearing up for this year’s Food Systems Summit as part of its broader effort to achieve the SDGs by 2030, it is imperative that the international community prioritize finding ways to conduct censuses safely. Most of the developing world is unlikely to have meaningful vaccination coverage until late 2022 or 2023. The UN has made efforts to mitigate the pandemic’s effects on agricultural censuses, recommending using existing administrative records and online surveys to source data. Yet with nearly three-quarters of people in rural sub-Saharan Africa lacking electricity, these efforts are no substitute for in-person census activities.

Quote from UN Secretary-General Antonio Guterres on the importance of the Pre-Summit

Despite last year’s global disruptions in data collection, there exist relatively simple solutions. By establishing safety protocols for in-person census taking and providing strong logistical and financial assistance to developing countries, the international community could help states regain key statistical capacity swiftly. What both developing and developed countries do in the coming months may determine whether 2020 was an outlier or an inflection point for global efforts to strengthen fragile states.

—Michael Harsch, Visiting Scholar, Weatherhead Center for International Affairs, and Alexandra Norris, BA Candidate, Harvard College
Michael Harsch is a Visiting Scholar with the Weatherhead Scholars Program. He is also a visiting assistant professor of international relations at the Frederick S. Pardee School of Global Studies, Boston University. His research focuses on international security, international organizations, conflict, and development. He is completing a book on “Islands of Stability in Fragile States.”
Alexandra Norris is a BA candidate in the Department of Government at Harvard College and is a research assistant for the project on “Islands of Stability in Fragile States” led by Professor Harsch.


  1. Sheep being gathered for the livestock census. November 7, 2007, Beïnam, Niger. Credit: FAO/Ado Youssouf

  2. 2015 Agricultural Census Advert. Credit: Statistics Botswana, YouTube

  3. State Capacity across the Developing World, 2005–2020. Note: World Bank Data on Statistical Capacity 2020; analysis by authors. A small number of coding errors were corrected (see documentation). Statistical capacity scores indicate whether a country lacks up-to-date versions of the following three types of essential records: agriculture census, population census, vital registry. Credit: Michael Harsch and Alexandra Norris, Github

  4. Farmer with traditional potato variety. March 27, 2014, Papua New Guinea. The country’s last agricultural census was conducted in 1961/1962. Credit: Bioversity International/P.Quek, Flickr (CC BY-NC-ND 2.0)

  5. Quote from UN Secretary-General, António Guterres. In 2021, Guterres will convene the UN Food Systems Summit to raise global awareness and promote global commitments and actions that transform food systems. The Secretary General is calling for collective action of all citizens to radically change the way we produce, process, and consume food. Accessed on May 18, 2021. Credit: UNFSS Communications Hub, Trello, UN Food Systems Summit 2021