Using experimental economics to study subsistence culture in arctic communities

Subsistence harvests are a primary food source for people living in many remote communities-but it is not without risks. Food may be plentiful in some years and scarce in others; with climate change, there is also the possibility that some subsistence resources might become more variable and less predictable in the future. Moreover, hunting and fishing success can be inconsistent and these activities can often be dangerous (e.g. seal hunting during spring break-up). How subsistence-dependent communities in Western Alaska and Kamchatka, Russia, cope with and adapt to risk is the focus of an interdisciplinary study by UAA economics professors Lance Howe and Jim Murphy.

Jim Murphy (far right) with villagers in the Kamchatka region of Russia who shared their
subsistence way of life with UAA researchers.

Since 2007, Howe and Murphy have been collaborating with anthropologists Dr. Colin West (University of North Carolina) and Dr. Drew Gerkey (University of Washington) on a project that combines methods in anthropology and experimental economics: findings from ethnographic surveys were used to design economic experiments to test hypotheses important in both economics and anthropology. In their National Science Foundation-funded study, "Salmon Harvests in Arctic Communities: Local Institution, Risk and Resilience," Howe and Murphy have focused primarily on how individual-level risk affects sharing and harvesting in the Kuskokwim region of Alaska and in the Kamchatka region of Russia (two locations very similar to each other in their subsistence culture and lifestyle but very different in terms of cash income and political institutions). More specifically, their recent fieldwork focuses on how individuals decide to share with other community members should someone face an unexpected setback or "shock."

Lance Howe in the field in the Kuskokwim region of Alaska.

Murphy explains that there are two different types of risk individuals could face: community-level risk and individual-level risk. "Community-level risk is what happens if the salmon don't come; it affects everybody," he says. "Individual-level risk is what would happen if you were sick or injured and can't go fishing."

In designing and carrying out the research, the team collaborated with traditional councils in six different Yup'ik and Cup'ik Western Alaska communities and three indigenous Koryak communities in Kamchatka, Russia. In Kamchatka, the research team engaged local people in a "public goods" decision-making experiment. In these exercises, participants decide how much of their time to divide between collaborating with the other four members of their group and how much time to spend working individually. If they choose to work individually, they get paid 10 rubles an hour, but if they collaborate with each other, the group gets paid an hourly wage of 20 rubles. It might seem obvious that everyone would choose to spend time working in the group activity, but there is a catch, all group earnings are divided evenly among the whole group-regardless of individual effort-consequently, there is a strong incentive to free-ride off the efforts of others. Over time, theory predicts most participants will allocate all their time to the private activity because of the free rider problem. This first set of decisions provides some insights into the conditions under which group members are able to sustain cooperation.

To this baseline exercise the researchers add environmental risk. They test how risk influences time allocation decisions and, more importantly, they test if groups insure one another against risk given the opportunity. For example, in this variation of the exercise or "treatment," one group member receives a "shock" each round, which means that they have unexpectedly lost all their money. However, other group members are given an opportunity to anonymously "share" their earnings with the player receiving a shock in this treatment. Similar to real life, this puts individuals in a position where they don't have the resources they need unless the rest of the community shares with them.

This sets up a second set of decisions the researchers are interested in: to what extent do participants condition the amount shared with the shocked player on the past sharing or group investment decisions of the player receiving the shock? Player names and associated activities always remain private, but in some variations of the exercise other group members know how much the player receiving the shock shared in the last period or contributed to the group activity.

"In this way we can enable participants to condition their decision based on knowledge of how much the shock victim contributed," explains Murphy. "Overall, the public goods game helps us try to tease out how much of the choice to share was conditional, and how much was unconditional, based on others' cooperation and sharing behavior in return."

Subsistence fish camps on Russia's Kamchatka Peninsula (top) are very similar to those
in Western Alaska (bottom).

The research team completed field experiments in Kamchatka last summer and just finished field experiments in Western Alaska this May. Based on preliminary analysis of the Kamchatka data they find that there are strong levels of sharing, or risk pooling, across all treatments. Given a risky environment, participants voluntarily share significant amounts of money with the player receiving a shock. This is an interesting finding because economic theory predicts, given the design of the exercise, that there shouldn't be high levels of sharing. At the same time, however, as more information is provided about the shocked player (such as how much they shared with others in the past or how much they contributed to the group activity), other players condition almost all their sharing on this information. That is, many participants choose not to help the shock victim if they haven't been generous to others in the past or if they have not contributed to the group activity.

The researchers are also working to resolve some puzzling outcomes. For instance, participants demonstrate high levels of cooperation in the sharing dimension of the exercise but very low levels of cooperation in the group investment decision. "In the Russian data," Murphy says, "it seems like the sharing decision and the cooperation decision are two completely different things and aren't tied together. In Alaska, we added another layer to the research and recorded all the conversations people were having in order to analyze the decision-making process, not just the final decision itself."

The current study is really just one small part of a much larger field of study. Firstly, their current study is only scratching the surface of individual-level risk and sharing norms that influence community-level resilience. Secondly, there are other questions to be answered related to aggregate shocks at the community level (for instance the collapse of a salmon run) as well as varied rules and norms for different food sources, or differences in norms between sharing food versus sharing cash.

"The way we would cooperate in moose hunting is different than that when fishing," says Murphy. "The rules and norms that worked well with sharing fish, might not carry over well to hunting larger game. How adaptive are those rules? How good are they at shifting from one food source to another?"

This summer, Howe and Murphy are wrapping up their individual-level field experiments in a number of villages in the Kuskokwim region of Alaska (including Chevak, Tuntutuliak, Nunapitchuk, Lower Kalskag and Upper Kalskag) and beginning to compare those findings with the data from Kamchatka (which includes the three villages of Tymlat, Ossora and Karaga).

Over the last five years, roughly 20 UAA students have participated as research assistants on the project, including Cristina Gaina, who traveled to Kamchatka to help conduct field experiments, and Dan Allen and Chloe Tanaka who helped conduct the field experiments in Western Alaska communities this year. Drs. Howe, Murphy and Gerkey will be returning to Kamchatka this summer with two more UAA students, Josephine Hishon and Mikhail Kolodiy. The purpose of this visit will be to provide feedback to the communities that participated in the study.

A full project abstract can be found at www.nsf.gov/awardsearch under Award #0729063.