A new statistical model developed by Stanford researcher Bill Burke helps identify barriers that keep farmers in Kenya from participating in the dairy industry. Results published in the American Journal of Agricultural Economics suggest that better technology, infrastructure, and access to markets may lead more people to take up dairy farming, which could improve rural food security and livelihoods.
 

Growing the dairy sector

Kenya’s dairy farms house 3.4 million cows, about 85 percent of all dairy cattle in East Africa. Yet within Kenya, the popularity of dairy products has risen steadily in recent years, and in 2006 the country became a net importer as demand outpaced supply.

Small family farms contribute the majority of Kenya’s dairy output, in addition to providing many rural jobs. The government of Kenya has targeted small-scale dairy operations as a key sector for boosting both domestic food supply and rural incomes.

Image

“A new government policy could aim to do three things,” said lead author Dr. Bill Burke, a research scholar at the Center on Food Security and the Environment at Stanford. “First, to increase the number of farmers who produce dairy, second, increase the number of those producers that participate in the dairy market, and third, to increase each farmer’s productivity and sellable surplus. And to design a good policy, you need good data and good analysis.”

Previous studies fall short

Previous studies of the Kenyan dairy market rely on a “double hurdle” model, a type of statistical analysis that measures two things: why producers become net buyers versus net sellers of a certain product, and how much of that product they ultimately buy or sell.

But these models assume that all farmers produce the same product, which is not the case in the dairy sector, said Burke.

Image

Dairy farming is labor intensive and has high overhead costs. Farmers without access to cold storage facilities, or without a nearby place to sell their produce, risk losing their inventory to spoilage. Even more risk comes from the fact that the weather and the price of cattle feed can both change suddenly.

“Dairy farming isn’t like growing rice or corn. It is expensive, specialized and risky, so not all farmers do it,” said Burke. “If we really want to understand how to grow the Kenyan dairy sector, we need to understand why some farmers choose to produce dairy in the first place, and why others don’t.”

A new level of analysis 

To better understand why some farmers participate in the dairy market, Burke and his co-authors, Robert Myers and Thom Jayne from Michigan State University, used data from 1,275 farming households collected by staff at the Tegemeo Institute of Agricultural Policy and Development, who visited each home four times between 1997 and 2007.

The team then designed a model to analyze market participation among farmers. They developed a new “triple hurdle” model, adding a new level of analysis to the double hurdle model that allowed them to identify the factors that influence some farmers and not others to produce dairy.  

To produce or not produce?

The team’s triple hurdle model offered a clear advantage over previous studies. “The new model results are significantly different than if we had used a double hurdle approach, and the magnitude of our estimates are a statistically better fit to the data,” said Burke. “It provided us with a more nuanced understanding of how policy can influence dairy supply.”

Burke’s team found that farmers were more likely to be dairy producers the wealthier they were in terms of land, household assets and credit. Farmers were also more likely to join the dairy sector when they had access to improved technology such as specially-bred “grade” dairy cows and “zero grazing” systems for cattle feeding. Another important factor was the presence of commercial dairy processors and informal traders, because they provided farmers with more certainty that their produce would reach buyers.

Image

These same factors – wealth, technology, and marketing channels - helped determine the market position of a dairy producer: whether they were net sellers or net buyers of dairy. Infrastructure also played an important role. Producers were more likely to be net sellers when they had a formal education and nearby access to electricity and drivable roads.

“These data, and the new insights from our triple hurdle model, suggest that market channels are important incentives for farmers to participate in dairy production,” said Burke. “But we also see a clear case for increasing government investments in infrastructure like roads, rural electrification, and education. We hope our research will inform the government of Kenya as they design a new policy approach to increase dairy farmer participation and productivity.”

CONTACT

Dr. Bill Burke: burkewi2@stanford.edu, (650) 724-1290

Laura Seaman, Communications Manager: (650) 723-4920

A July 2014 research paper co-authored by FSE deputy director David Lobell is one of 25 articles selected by the editors of the scientific journal Environmental Research Letters to be featured in the journal's Highlights of 2014 collection. The Editorial Board also recognized the paper as 'Highly Commended' during the vote for ERL's 'Best Article' for 2014.

In "Getting caught with our plants down: the risk of a global crop yield slowdown from climate trends in the next two decades," Lobell and co-author Claudia Tebaldi, a senior researcher at the National Center for Atmospheric Research, find that the chance of a worldwide slowdown in yield growth for wheat and corn in the next 20 years is significantly higher due to global warming. With only natural climate variability the odds of a 10 percent yield drop over two decades - equivalent to a halving of the yield growth rate - are one in two hundred for wheat and corn. But under a scenario of human-induced climate change, the likelihood rises to one in ten for corn and one in twenty for wheat.

Lobell said these results are of particular interest to organizations working toward global food security and related issues that would be strongly impacted by falling crop yields in the next 20 years. 

“The people asking these questions are accustomed to planning for scenarios with much less than a 10 percent chance of happening, so it will be interesting to see whether this study has any effect on how they operate," Lobell said. “As scientists, we might prefer to work on time scales in which the answers are clearer, but we also want to be responsive to the actual concerns and questions that decision makers have.”

Lobell is associate professor of Environmental Earth System Science at Stanford and associate director of the Center on Food Security and the Environment. He is also a senior fellow at theStanford Woods Institute for the Environment and the Freeman Spogli Institute for International Studies.

The European Union led the world in wheat production and exports in 2014-15. Yet Europe is also the region where productivity has slowed the most. Yields of major crops have not increased as much as would be expected over the past 20 years, based on past productivity increases and innovations in agriculture.

Finding the causes of that stagnation is key to understanding the trajectory of the global food supply.

Logically, it would seem that climate change would affect crops. But in the overall picture of agriculture, it's hard to figure out how much. European farming is a complex venture, and other possible stagnating factors include changes in government policy. For example, farm subsidies are no longer based on productivity and the use of fertilizer is now controlled to reduce runoff into water supplies. Ongoing positive factors include improvements in farm management practices and advances in crop genetics.

Historically, scientists relied on models to estimate the effects of climate change. Now Stanford's Frances C. Moore has for the first time statistically quantified the relative importance of climate in the stagnation of European crops. She found that warming and precipitation trends are affecting European grain harvests. Moore is a PhD candidate in the Emmett Interdisciplinary Program in Environment and Resources.

"This study is sobering in that it shows climate drags on some of the crops in this region," said David Lobell, co-author of the paper. "Yet this new approach to looking at the problem will help us understand more quickly what impacts require more attention, and that can only be positive in the long term." Lobell is an associate professor of environmental Earth system science and the deputy director of the Center on Food Security and the Environment at Stanford. He is also a senior fellow in the Freeman Spogli Institute for International Studies and the Stanford Woods Institute for the Environment. He studies ways to improve crop yields in major agricultural regions, with emphasis on adaptation to climate change.

"This is a major step in using quantitative analysis to disentangle the effect of climate change in a complicated system," said Dáithí Stone, a pioneer in comparing actual seasonal weather forecasts with what those forecasts would have been if human activities had not emitted greenhouse gases. "It demonstrates that the signal has become large enough that we may see the effect of climate change in a complicated system like agriculture." Stone is a research scientist in the Computational Chemistry, Materials and Climate Group of Berkeley Lab.

How wheat and corn and barley grow

Moore considered two factors in the study: actual crop yields and expected crop yields given historic temperature and precipitation trends. She applied statistical analyses to look for patterns in regional maps of actual European yields of wheat, maize (known in the United States as corn), barley and sugar beets, from 1989 to 2009.

The study found that climate trends can explain 10 percent of the slowdown in wheat and barley yields, with changes in government policy and agriculture likely responsible for the remainder of the stagnation. Moore found evidence that long-term temperature and precipitation trends since 1989 reduced overall European yields of wheat by 2.5 percent and barley by 3.8 percent, while slightly increasing maize and sugar beet yields.

Moore also wanted to find out to what extent farmers had adapted their practices to accommodate changing conditions. She applied power analysis, a statistical tool to test the effect of adaptation. But she discovered the test was not effective in the context of this study.

"We think farmers have been hurt already by warming and drying trends in Italy," Moore said. Undaunted by the limits of statistical analysis to measure farmer adaptation, she is planning another way to find out. "I have been doing this work in front of a computer – in the future I would like to go to Italy," she said. "It would be interesting to talk to the farmers."

Leslie Willoughby is an intern at Stanford News Service.

Media Contact

Frances C. Moore, Emmett Interdisciplinary Program in Environment and Resources: (617) 233-3380, fcmoore@stanford.edu

Dan Stober, Stanford News Service: (650) 721-6965, dstober@stanford.edu