Food assistance programs and food insecurity: implications for Canada in light of the mixing problem

Craig Gundersen, Brent Kreider, John Pepper, Valerie Tarasuk

Research output: Contribution to journalArticlepeer-review


In light of concerns about high rates of food insecurity, some have suggested that it might be time for Canada to implement national food assistance programs like those provided in the US, namely the Supplemental Nutrition Assistance Program (SNAP) and the National School Lunch Program (NSLP). In this paper, we assess how adopting these types of assistance programs would change the food insecurity rate in Canada among households with children. Using data from the Current Population Survey (CPS), we first evaluate the causal impact of these programs on food insecurity rates in the US using the Canadian definition of food security. Following other recent evaluations of food assistance programs, we use partial identification methods to address the selection problem that arises because the decision to take up the program is not random. We then combine these estimated impacts for the US with data from the Canadian Community Health Survey (CCHS) to predict how SNAP and NSLP would impact food insecurity rates in Canada. Partial identification methods are used to address the “mixing problem” that arises if some eligible Canadian households would participate in SNAP and others would not. The strength of the conclusions depends on the strength of the identifying assumptions. Under the weakest assumptions, we cannot determine whether food insecurity rates would rise or fall. Under our strongest nonparametric assumptions, we find that food insecurity would fall by at least 16% if SNAP were implemented and 11% if NSLP were implemented.

Original languageEnglish (US)
Pages (from-to)1065-1087
Number of pages23
JournalEmpirical Economics
Issue number3
StatePublished - May 1 2017


  • Food insecurity
  • Mixing problem
  • National School Lunch Program
  • Nonparametric bounds
  • Partial identification
  • Selection problem
  • Supplemental Nutrition Assistance Program
  • Treatment effects

ASJC Scopus subject areas

  • Statistics and Probability
  • Mathematics (miscellaneous)
  • Social Sciences (miscellaneous)
  • Economics and Econometrics


Dive into the research topics of 'Food assistance programs and food insecurity: implications for Canada in light of the mixing problem'. Together they form a unique fingerprint.

Cite this