Table 2. Estimated rebate effects on household food purchases at designated supermarkets.

Dependent variable	All shoppers		Nearby shoppers
	10% Rebate	25% Rebate	10% Rebate	25% Rebate
Ratio of healthy to total food expenditure	0.0127** (0.0008)	0.0195** (0.0008)	0.0182** (0.0032)	0.0254** (0.0032)
Ratio of fruit/vegetable to total food expenditure	0.0057** (0.0006)	0.0085** (0.0006)	0.0063* (0.0022)	0.0091** (0.0022)
Ratio of less-desirable to total food expenditure	−0.0106** (0.0009)	−0.0137** (0.0009)	−0.0145** (0.0032)	−0.0165** (0.0032)

Note: Boldface indicates significance. Reported parameters (SE in parentheses) are estimated using linear household fixed-effect models, controlling for month/year fixed effects. Nearby shoppers (n=56,908 households) are defined as those who lived at least 1 kilometer closer to the nearest Pick n Pay supermarket relative to the nearest Shoprite or Woolworth's supermarket. The Eicker-Huber-White sandwich estimator is used to calculate SEs clustered at the household level.

^*p<0.01,

^**p<0.001