When Is a Drought Over? Herding Families in Northern Kenya Recovered on Different Clocks

Researchers measured the same herding communities before, during, and after one of the region's most severe droughts in recent history. Reported hunger and thirst were back near their old levels by mid-2024, a little over a year after the rains returned. The herds, the children's growth, and the old patterns of travel did not.

The rains came back to Illeret in 2023. Illeret is a cluster of settlements on the northeastern shore of Lake Turkana, close to where Kenya meets Ethiopia, and it had gone through five failed rainy seasons in a row, beginning in late 2020. The long rains arrived in March. The researchers who later studied the area date the drought's end to that month, and aid agencies, as the researchers put it, declared the drought over.

In June and July of that year, a research team sat down with Daasanach families around Illeret and asked them about water. Seventy-one percent of the people surveyed gave answers that placed them in the most serious category of water insecurity. The year before, in the middle of the drought, it had been about 41 percent. Before the drought, about 34.

Hunger went up too. The same rains that ended the drought also brought floods, and the floods blocked food aid from reaching the area for more than three months. The team points to those missed deliveries as the likely reason food insecurity got worse after the rain rather than during the dry years, though they didn't test that separately. Nothing in the paper says the declaration slowed the aid down; the floods did that on their own. What the timing does show is how little a rainfall-based declaration told anyone about life inside these households. An outside institution said the crisis had ended in March, and in July families were going without more than they had at the height of it.

A study published this year in the journal Global Environmental Change, led by Penn State anthropologist Asher Rosinger, is built around that mismatch. It asks something that sounds easy to answer: when is a drought over? Going by the families' data, it ends at different times for different parts of a life, and some of those parts still hadn't come back when the study stopped counting.

Measuring before anyone knew a drought was coming

The Daasanach are herders. Their households have long been organized around goats, sheep, and cattle: moving animals to water and grass, drinking their milk, trading them, giving them at marriages. In the dry country around Lake Turkana, that way of life is built to handle scarcity, though five failed seasons back to back went well beyond the hard years it was built for.

Rosinger's team first measured these communities in mid-2019, with no idea what was coming. "We didn't know we were about to witness one of the most severe droughts in recent history," he told Penn State.

That accident of timing is what sets the study apart. Disaster research almost never has a "before," and without a record of where people started, researchers can't say how far they fell or whether they've made it back.

The team did four rounds of surveys and measurements, each in June and July: 2019; 2022, deep in the drought; 2023, just after the rains returned; and 2024, roughly a year after the drought ended. The pandemic kept them away in the years between. One Penn State photo from mid-2022 shows Rosinger using calipers to pinch a fold of skin on a man's back, just under the shoulder blade. Skinfold thickness gives a rough reading of stored body fat, the reserve fuel a body draws on when food runs short. By the following round, adults' body mass had dipped.

Over all the rounds, the researchers surveyed and measured more than 500 adults and nearly 400 children.

Measuring what going without feels like

Two of the study's main tools are short questionnaires. The water survey asks 12 questions about the past four weeks: whether you worried about water, whether there was too little to wash your hands or your clothes, whether you went to sleep thirsty, whether you felt ashamed of your water situation. The food survey asks nine along the same lines: whether you ate foods you didn't want, ate smaller or fewer meals, went to bed hungry, went a whole day and night without eating. Each adult head of household was interviewed separately, so a husband and wife could describe the same home's water differently.

Neither survey measures rainfall or calories. They ask about what scarcity was like and what people had to do to get through it, and that's why I trust them as a reading of daily life.

Food and water didn't move in lockstep. Water insecurity had already crept up by 2022. Food insecurity showed no clear change that year, which the researchers put down to aid that was still getting through. Both rose sharply in 2023. By 2024 both had come back down, and serious water insecurity was at 27 percent, a little below where it started.

The researchers flag one catch themselves: 2019 was already a fairly dry year. "Back to baseline" means back to a year that wasn't easy either.

With only these two surveys in hand, you'd call the drought over by 2024. The rest of what the team measured says otherwise.

Herds still far below where they started

In 2019 the average household had 24.6 sheep and goats. In 2023 it had 8.0. By 2024 that was back up to 13.9, still about 44 percent short of where families began.

Cattle fell further and recovered less. Households averaged 6.2 head in 2019, 1.0 in 2023, and 1.5 in 2024, about three-quarters below the starting point.

A herd is food and savings, and it's also breeding stock. When an animal dies, so does every animal it would have gone on to produce. My grandmother farmed maize in the Guatemalan highlands before my family came north, and when she talked about the bad years, the worry that stayed with her longest was how much seed had been saved. The next harvest could only be as big as what went into the ground. Herds work the same way. Goats breed younger and more often than cattle, so a goat herd can rebound within a few seasons, while cattle take years. You can see that difference plainly in the 2024 numbers.

Herders have never rebuilt by waiting on births alone. The anthropologist Toru Sagawa, who has spent years with the Daasanach, describes bond friendships through which people trade crops and small animals, sometimes with neighboring groups the Daasanach otherwise treat as enemies. Some of these friendships start when one person names another's child, and they can bring a share of bridewealth, the animals one family gives another when a couple marries. Sagawa is firm that these ties are about hospitality and real affection, and that treating them only as insurance misreads them.

Work among the neighboring Turkana people points in the same direction. Families there rebuilt herds through marriage payments, gifts, and trades, and kept partners who lived farther off or grazed different land, so that one family's bad year was less likely to land on the other at the same time. Recovery worked when those partners still had animals to give and stalled when they'd run out.

World Weather Attribution, a group that estimates how much climate change shaped particular weather events, calculated that human-caused warming made a drought this severe in the southern Horn of Africa about 100 times more likely. The figure covers low rainfall together with extra drying from heat, and the group found no clear trend in the rainfall itself. Much of the severity came from higher temperatures pulling more water out of soil and plants, so the same rain left the land drier than it used to.

A drought that big covers a whole region at once, which is hard on a system that depends on having friends far away. The partner you'd ask for goats may have lost theirs too.

The Kenya study didn't measure these friendships or trades, so it can't say how much of the goat rebound came from births and how much from gifts, trades, or purchases, and the Turkana research is about neighbors, not these families. What the study does show is that however families were rebuilding, they were still climbing out of a deep hole when it ended.

Children still behind a year later

Researchers track children's growth by comparing each child with an international reference for healthy children of the same age and sex. A falling score means a child is putting on weight or height more slowly than that reference. It doesn't mean the child is shrinking.

Children's weight for their age fell in 2022 and fell again in 2023. In 2024 it was still a little lower than in 2019, though by then the gap was small enough that chance couldn't be ruled out. Height for age had clearly dropped by 2022, and the paper reports it hadn't fully recovered by 2024. Body size relative to age, and upper-arm circumference (the tape health workers wrap around a child's arm to screen for malnutrition), both got worse through 2023 and only partly recovered in 2024.

Milk is at the center of a herding family's diet, and children were drinking far less of it in 2022. After that it gets murky: in one place the paper says children's milk consumption went back to normal, and in another that it stayed lower.

Adults bounced back faster. Their average body size dipped in 2023 and partly recovered, and by 2024 their body fat was slightly higher than in 2019.

One caution belongs here. An earlier study of Daasanach children found that many fall behind the international reference when they're young and then catch up a good deal in height later, and its authors note that the reference doesn't fully account for normal differences between populations. So this study can't tell us these children will grow up shorter. It can tell us that a year after the rains came back, the adults had mostly recovered and the children were still behind on several measures.

More thoughts of moving, fewer actual moves

The survey question I keep returning to is a short one:

The team's 2024 paper asked it in 2019 and again in 2022. The share of people who had thought about moving at least once went from 55 percent to 92 percent. Over the same two rounds, the average number of moves people actually made went down, from 4.2 to 3.3.

The new paper follows the moves further, through 2023 and 2024, with a somewhat different group of participants. It counts a "move" as any stretch away from home longer than a week in the past year, leaving out market trips and social visits, so it mostly picks up herding journeys and seasonal travel rather than permanent relocation. People averaged 3.9 such moves in 2019, 2.2 in 2023, and 2.3 in 2024. The 2024 drop falls just short of the usual bar for statistical confidence, but there's no sign of it climbing back to the old level. The sections of the new paper I could see don't report a later figure for how often people thought about moving, so whether that worry eased after the rains is still an open question.

For herders in dry country, moving animals toward water and grass has long been the main way through a hard season. The paper's title says the drought pushed diet and movement "away from traditional pastoralism," pastoralism being the herding way of life.

The researchers offer a few reasons people moved less. Smaller herds need fewer people traveling with them. Families gathered near Illeret, where there are water points, schools, and services. And there may simply have been no better grazing to go to. The survey didn't ask people why, so these are the researchers' interpretations.

People who study migration would read this pattern more than one way. A study of rural households in Thailand and Vietnam found that two consecutive years of moderate drought lowered the share of households sending someone away by 5.3 percentage points, and the authors point to families having less money and fewer assets to pay for a move. Kerilyn Schewel, in a widely cited review, argues that wanting to move and being able to move are separate questions, and that the people who stay include those who choose to, those who can't leave, and those who neither want to nor could.

The Daasanach numbers fit both readings. A family with 8 goats instead of 25 has less reason to travel with a herd and fewer resources to travel at all. A family that settled near a water point and a school may have decided staying is the better life. Plenty are probably weighing both at once. The two measures also cover different stretches of time, since thoughts about moving were asked about the past month and moves were counted over the past year. Whether staying put was forced, chosen, or the beginning of a different way of living is something the families are best placed to answer, and the surveys haven't asked them.

A once-looked-down-on food on more plates

During the drought, more people started eating fish from the lake. The share of adults who'd eaten any fish in the past week went from about 39 percent in 2019 to about 63 percent in 2022. By 2024, adults were eating fish about 0.6 more days a week than at the start, and eating less meat.

Fish carries social weight for the Daasanach. The paper describes the Dies, a Daasanach term for poorer community members who don't own cattle and are associated with fishing. A related study from the same project, a preprint that hasn't finished review by other scientists, summarizes older research showing that herders sometimes accepted fish from Dies friends partly because turning down hospitality would be an insult, even while they looked down on fish as food.

Attitudes had started shifting before this drought, at least in parts of Daasanach country. In a university interview, Sagawa described fishing camps on the Ethiopian side of Daasanach land by 2015, including young people who still owned animals and new traders buying fish. People there felt fishing sat alongside herding better than plantation work did. The drought may have sped up a change already happening elsewhere among the Daasanach, though no one has traced that timeline in the Kenyan villages.

In the preprint's group discussions, people talked about fish as food, as income, and as a way to buy things that last. How they felt about it depended on how far from the lake they lived. They also named what stood in the way: getting nets and boats, getting fish to market, crocodiles, and illness linked to the water. These come from conversations, not measured rates, but they're the closest this research gets to the community explaining its own choices.

The lake itself depends on decisions made somewhere else. Lake Turkana's plants, animals, and fisheries rely heavily on the Omo River, which flows in from Ethiopia, and on the timing of its floods. A satellite study linked shifts in the river's flow, tied to upstream dams, to changes in the lake's algae, the base of its food chain. No one has shown this affected fishing near Illeret during the drought. Still, families are leaning harder on a food source whose water is managed by another government.

In the families' telling, fish is a workable adjustment with real dangers attached. Whether it also feels like a loss, set against the herds it partly replaced, the discussions as summarized don't say.

The people the study lost track of

Long studies lose people, and this one lost a lot of them. A quarter of the adults measured in 2019 (25.5 percent) never took part again, and neither did 43.1 percent of the children. Only 51 adults were measured in all four rounds. The team brought in similar neighboring households to fill the gaps, which keeps the communities represented but tells us nothing about what happened to the people who dropped out. The researchers acknowledge those people may be different from the ones who stayed.

That affects every recovery number above. If the families who fell out of the study were the ones who traveled farthest, or the ones who lost the most, the averages would look different with them in, and the study has no way of knowing which.

The researchers list other limits as well: a single "before" measurement, a dry baseline year, and no comparison group of Daasanach who were spared the drought.

Questions to ask when something is declared over

What people said about hunger and thirst, the animals they owned, how their children were growing, and how far they traveled each recovered at its own pace. Biology set part of that pace, and whether help could get through the floods set another. The study didn't measure how much relationships between families mattered.

The study doesn't tell anyone how to recover. It does raise questions worth asking after a freeze, a fire, a flood, or anything else that gets declared over:

• Who said it was over, and what were they measuring when they said it?
• Which losses can only be rebuilt as fast as things grow, and who could help speed that up, if they weren't hit at the same time?
• What did you stop doing during the crisis that you haven't picked back up, and do you know whether that's because you can't or because you've decided not to?
• What did you start doing that you once looked down on, and what does it mean to you now?
• Who from before is no longer around to be counted?

The paper notes that the team went back after the final survey to watch and talk with families. The published numbers stop in 2024. Around Illeret, people are still working out how many animals they can afford to keep, how far to travel with them, and how much of their food will come from the lake.