Munger and Roberts have a great Econtalk where they briefly discuss restaurant pricing schemes and the puzzle of why it is not usually cheaper to have take-away (a.k.a. "to go") when clearly the cost of producing the good is cheaper than if one doesn't eat-in. Having lived in London for nearly four years, I thought I might provide some relevant observations from this side of the pond. Here, many restaurants do charge lower prices for take-away than for the same item eaten inside the restaurant. Here is one example from my local Starbucks:
One hypothesis is that this is more common than in the US because of the high real estate prices in London. Due to just how expensive land is here, the price of taking up space is better accounted for by restaurants. I think this helps explain the pervasive practice of setting time limits on tables in restaurants in London, but not the price differential between eat-in and take-away. This pricing strategy is at work in other areas of England where real estate prices are much lower. And we also don't seem to see the price differential emerge in high priced real estate US cities like San Francisco. So, that doesn't seem a good answer.
Perhaps the answer lies in the divergent tipping norms in these two countries. In England, for many restaurants, it is not expected that guests tip at all. If the meals for take-away and eat-in were the same price, this would be an example of divergence of price from marginal cost. Thus, restaurants price take-away lower than eat-in because it is cheaper to produce. In the US, by contrast, tipping norms mean that you are paying less for take-away rather than eat-in. This is consistent with the idea that the tipping norms affect "official" prices because if tipping doesn't exist, then the restaurant owner must explicitly and intentionally discount the prices.
A corollary prediction, supported by my casual observations, is that in those London restaurants that do include a "recommended service charge" for eat-in service (which is usually 12.5%), these are the restaurants that do not have price differentials on take-away and eat-in.
Finally, my brilliant husband likes to point out the absurdity of the oft-made claim (see here, here, and here for example) that tipping should be done away with because most people give the same tip regardless of the quality of the service. We're both service industry veterans, but also economists, so can attest to the fact that what matters is not the average responsiveness of tippers, but the marginal ones. The average response is irrelevant. The marginal consumers who do adjust their tips based on quality can and do send a clear message to reward quality and punish poor service.
Addendum: my colleague John Meadowcroft points out that the VAT in the UK is 20% on eat-in and lower on take-away. The rules differ based on whether the food is hot or cold. Cold items can take advantage of the lower VAT, and thus the lower take away price, but hot items are subject to the higher VAT. A twitter exchange with a popular chain in the UK confirms this practice within their branches (and they report that cold food is actually "zero rated" for VAT).
This is anecdotally confirmed with closer inspection of my local Starbucks. The "luxury fruit bread" is served warm and toasted. The cinnamon swirl is not. And the eat-in price of the cinnamon swirl is exactly 20% higher than the take-away price.
It's compelling that the price difference in the UK is mostly the result of the VAT. As such, a great example of how taxes increase the cost of production and prices reflect the those marginal costs. But doesn't help to explain why you don't see different eat-in take away prices in more US restaurants.