Subsidies for store-bought foods may crowd out the consumption and production of country foods obtained through hunting, fishing and trapping by lowering the relative price of market foods and shifting demand toward store-bought alternatives. Yet country foods are nutritionally and culturally significant for Indigenous peoples, providing essential nutrients and sustaining cultural continuity, land-based knowledge and sharing networks. Policies that affect harvesting could have implications not only for food security but also for Indigenous food sovereignty, understood as the ability to maintain culturally valued food practices and autonomy over food systems. This paper studies how subsidies for store-bought foods affect participation in country food harvesting, leveraging variation generated by the replacement of the federal Food Mail Program with Nutrition North Canada (NNC). Using a difference-in-differences approach, I find that NNC eligibility reduced the probability that an individual hunts, fishes or traps by 14 to 16 percentage points after the reform. The decline persists through 2022, with no evidence that its magnitude becomes larger despite higher subsidy intensity. Harvesting for community sharing declines less in treated communities than in the comparison group, yielding a positive 13 percentage point treatment effect in 2022. Joint multiple-testing inference provides no evidence of differences in treatment effects across policy periods, reasons for harvesting, gender and age groups. By examining how retail food subsidies influence food production choices in northern remote communities, this paper informs policy discussions on food security while highlighting that lower harvesting participation does not necessarily imply a corresponding decline in the role of harvesting in supporting community food access.

Following the 2011 introduction of Nutrition North Canada, eligibility for the retail food subsidy program was associated with a 14–16 percentage point decline in the probability that an Indigenous individual hunted, fished or trapped.

Traditional foods (often referred to as “country foods”) play a critical role in Indigenous culture, nutrition, food sovereignty and food security. Policies that prioritize store-bought food are often viewed as undermining these foods. We provide evidence that two recent policy reforms that subsidize store-bought food in the Northwest Territories do not appear to negatively impact traditional foods. Our analysis uses community surveys conducted in the Northwest Territories that measure the importance of consumption and production of traditional country food along several margins. Using a difference-in-differences methodology, we show that two transitions associated with the Nutrition North Canada subsidy program—the transition from Food Mail to Nutrition North from 2011 to 2012 and the expansion of subsidy eligibility to new communities in 2016—had a zero or positive impact on these outcomes. We use price and quantity data from multiple sources to provide evidence on mechanisms and find support for an interpretation based on country foods having a high income elasticity and low substitutability with store-bought meat and other foods.

Households in many rural economies obtain food through both market purchases and household production. How retail food subsidies and harvesting support shape household food allocation depends on how readily households shift between these sources. We study this question in predominantly Indigenous communities in northern Canada, where households combine store-bought market food with country food obtained through hunting, fishing and trapping. We develop a two-stage household production model in which households allocate resources across market food, country food production and non-harvesting time. We use differential subsidy increases across communities under the 2019 Nutrition North Canada (NNC) reform as an instrument for effective market food prices and use the resulting demand response to discipline the model's central substitution elasticity. Counterfactuals show that removing the retail food subsidy reduces market food consumption substantially more than country food consumption. Counterfactuals motivated by the NNC Harvesters Support Grant increase country food consumption, whether support raises harvesting productivity or lowers input costs. Harvesting support leaves households' response to subsidy removal largely unchanged. The results suggest that retail food subsidies and harvesting support operate through different margins of the food system, with the former primarily affecting market food consumption and the latter expanding nonmarket food production.

This paper develops a multiple testing procedure for comparing treatment-effect summary parameters in staggered difference-in-differences designs. The procedure uses a bootstrap-based stepdown approach to test whether individual parameters differ from zero and whether any two parameters differ, while asymptotically controlling the familywise error rate.  The procedure also admits a simple graphical representation. Monte Carlo simulations show good finite-sample control of the familywise error rate together with meaningful power gains relative to a single-step approach. An empirical application illustrates how the procedure complements existing methods.