Economists Recalculate the Cost-Benefit Accounting of Local Food

Economists are making the argument that local food is not more energy efficient and less harmful to the environment (e.g., in emitted carbons) than factory food distributed to consumers at great distances. Factory food is distributed wholesale, which is (or might be?) more energy efficient than consumers individually driving to local produce markets. This might be an illustration of the point, I suppose: food for 100,000 consumers could be delivered from Ohio to New York City on one freight train, using less fuel and emitting less carbons, than food obtained by 100,000 consumers driving their cars from New York City to surrounding farmers' markets. But has anyone actually done empirical studies to back up such accounting? For instance, in this illustration, 100 local farmers deliver the food for 100,000 NYC consumers into the Union Square farmers' market, and the 25,000 consumers buy food at the market for themselves and three other persons each; and they  reach the market by taking the subway or walking. Now, redo the accounting.