The real failure curve
Restaurants do fail at higher rates than typical small businesses, but not the "9 out of 10 in the first year" figure that gets repeated. CHD Expert data and BLS BED suggest 40% survive year one and 20% survive year five — bad, but materially better than urban legend.
Failure modes shift over the lifecycle: first-year closures are usually undercapitalization (cash runs out before traffic builds). Year 3–5 closures are typically lease cliff events (rent reset) or chef/operator turnover.
Why state matters
Utah, Vermont, and Idaho lead 5-year restaurant survival rates (25–28%). New York, California, and Nevada trail (14–18%). The difference is mostly competition intensity + commercial-rent inflation: dense urban markets with high rent reset more restaurants more often.