New alternatives, same orders: modern alternative proteins did not replace meat purchases in restaurants

Plant-based alternative proteins are widely expected to displace meat consumption and, in turn, reorient the global food system away from the externalities of animal agriculture, but little empirical work has assessed these presumed substitution effects in naturalistic settings. Here, we use a dataset of 1.2 million transactions across seven restaurants and eight years to investigate whether introducing modern alternative proteins to restaurant menus reduces purchases of meat-containing dishes. We analyzed effects on total meat purchases and on category-matched purchases, pairing each alternative protein with the meat it emulates; we also estimate individual (within-customer) purchasing changes. Our results from 40 preregistered outcome-exposure pairings are overwhelmingly null, whether pooled across restaurants or stratified by restaurant. These findings challenge the widespread belief that alternative proteins displace meat purchases.