Prashant Yadav

Matching Uncertain Demand and Supply for Health Technologies: There’s much to be learned from others, especially during holiday season

Each year the holiday season highlights the complex task of matching demand and supply for gifts, Christmas trees and Thanksgiving turkeys. If you live in the United States you are perhaps still recovering from the lingering of Thanksgiving turkey. According to the National Turkey Federation, more than 730 million pounds of turkey (approximately 45 million turkeys) were consumed in the United States during Thanksgiving this year. Grocery stores, where people buy turkey, plan up to six months in advance to ensure they have the right quantity in stock. They establish long-term contracts with turkey producers far in advance of the Thanksgiving season. As a result of the long-term contracts, the turkey producers (especially the smaller farms which cannot afford to invest in their own market research) can better understand any shifts in demand each season. While the system is not perfect, it gets most Americans a Thanksgiving turkey (if they want one) at a relatively stable price (often decreasing in an inflation-adjusted sense).

If you celebrate Christmas and decorate your home with a Christmas tree, matching demand and supply for Christmas trees is also quite a complex problem. Christmas trees must be grown, cut and shipped, all within a short time span. It takes almost five years for a seedling to grow into the size tree that is typically used in homes. Approximately 35 million Christmas trees are harvested each year in the United States. Here again, retailers engage in long-term contracts with tree growers. Christmas tree demand has been decreasing over the past six years, as younger people who live in large cities often don’t have space for a real tree and artificial trees have become more common. In addition to changing demand and long lead times of growing and harvesting, matching demand and supply for Christmas trees is complicated by yet another factor: supply uncertainty. Heavy snow early in the season hampers timely tree harvesting. Michigan (where I live) is the third largest producer of Christmas trees and early season snow is not uncommon. Such events lead to increases in wholesale spot market prices for trees, but prices at large retail stores don’t fluctuate as much; the retailers have long-term contracts with tree suppliers.

The global market of olive oil (not as much a holiday staple) in many ways exemplifies the complexities of matching demand with uncertain supply. Spain and Italy produce approximately 70 percent of the world’s olive oil. While the demand side of the market has been relatively stable in the past few years (global demand of approximately 3,000 tons), the supply side of olives is quite challenging. The olive oil industry is highly fragmented with millions of farmers supplying a few thousand oil extraction mills and a few hundred refineries, many of which operate below capacity. Olive production each year is also highly uncertain due to uncertain crop yield. For example, this year a drought in Andalusia, Spain, has almost halved the country’s olive production. A bacteria in Tuscany has significantly decreased Italian production as well. Prices are now rising. The fragmented nature of the industry and the wide swings in production due to uncontrollable events mean that long-term contracts cannot insulate the market from price and supply shocks.

Similar challenges with matching uncertain demand with uncertain supply exist in the market for malaria medicines. A shortage or even a small price increase in this market has far more significant consequences than not having a Christmas tree this year, or having to pay a higher price for Italian olive oil. Undoubtedly, matching demand and supply in this market is a much more serious matter; a shortage or higher prices can lead to death and greater spread of disease.

Artemisinin combination therapy (ACT), the most effective malaria medicine today, is manufactured from artemisinin which is extracted from the plant Artemisia annua. The entire cycle of growing Artemisia to packaging the tablets takes somewhere close to 14 months. The plant is grown by farmers in China and Vietnam and artemisinin is extracted from it by more than two dozen extractors. Patients obtain malaria medicines in government-run clinics or in private sector pharmacies and drug shops. International agencies, most notably the Global Fund to Fight AIDS, Tuberculosis and Malaria and the U.S. President’s Malaria Initiative, provide financing to developing country governments for purchasing ACTs. Predicting the demand for ACT remains challenging due to financing delays, changes in epidemiology and many other factors.

On the other hand, there is also a significant supply uncertainty in growing Artemisia annua. Farmers can switch to other crops and sub-optimal rains can hamper crop yield. Long-term contracts with ACT manufacturers were recommended as a strategy in 2006, and recently the Global Fund has initiated two-year contracts with ACT manufacturers. A soft loan program was set up for artemisinin extractors with the hope that it will incentivize more farmers to grow artemisia through pre-financing and long-term contracting. A synthetic form of artemisinin was launched this year that cuts short the long lead time of artemisinin cultivation and allows manufacturing the product in less than two months.

Several options exist to improve the matching of demand and supply in this market. Many have been discussed and a few have been tried. In a recent paper, co-authors Burak Kazaz, Scott Webster and I develop a model of demand and supply matching in the malaria medicines market using the “modeling machinery” we have used to study other markets with uncertain demand and uncertain supply. It turns out that increased forward contracting and better demand forecasting don’t yield as strong a benefit as improving average yield of agricultural artemisinin production, and creating a larger and carefully managed supply of semi-synthetic artemisinin.

Markets have great richness in context. Developing tools for improving markets to serve larger societal needs requires a deep understanding of the market context and tailor-made analysis. However, making comparisons with other markets and having close interactions with those who work on a completely different market can be immensely valuable. There is a high chance you will learn something that may apply to your particular market. Worst-case, you will become better at describing the uniqueness in the institutional context, market structure and technological stage of the market you focus on.

So as you walk to work tomorrow think of the newspaper kiosk or the woman who runs the flower shop at the corner and how they manage their demand uncertainty. Do they do long-term contracts? Have they found a new way to reduce risk in their market? Have they found a new way to share their risk with others? Do they get volume guarantees? Have they found new ways to reduce lead time? Don’t hesitate to ask. The worst that can happen is they’ll call you a market-analysis and supply-chain nerd. And, after all, it’s the holiday season!

Prashant Yadav is a senior research fellow at the William Davidson Institute (WDI) and director of the Health Care Research Initiative at WDI.

Health Care
William Davidson Institute