Do people spend more if they have a credit card? The answer will depend on who you ask.

Credit card companies say YES! when they extol the benefits of taking cards to grumpy merchants. And credit card bashers say YES! too when they pile on reasons for why credit cards should be banned. But, the reality is that evidence on this has been pretty weak. Sure, the average credit card user may spend more than the average cash customer, but this might be because they are simply different kinds of customers. It isn’t as if carrying a credit card made a well-heeled woman buy a fancy new outfit at Neiman’s; being well-heeled is the reason that woman was able to get and carry an Amex card, or a Neiman’s card.

Most studies of the effect of credit cards on purchase behavior aren’t really capable of showing that having a credit card causes a person to buy more. As it is well known there’s a big—and really important difference—between correlation (well-heeled women both carry cards and shop more) and causation (giving a card to someone who doesn’t have one, causes them to buy more). Fortunately, a new paper by two Carnegie Mellon behavioral economists provides some interesting evidence on causation and credit card spending.

These economists did a controlled experiment at the cafeteria of a large insurance company that enabled them to disentangle causation from correlation. The authors are firmly in the “Let’s Ban Credit Card Camp.” But, surprisingly to them, they discovered that there wasn’t much difference between people who used cash and people who would ordinarily pay with cash but were incented to pay with a credit card instead. The cash users spent $4.59 on lunch while the credit card users spent $4.93—a difference that wasn’t even close to statistical significance.

This experimental study of the effects of credit card use on spending is the first of its kind and surely won’t be the last. Maybe there was some hidden flaw in their methodology. Or maybe using a card doesn’t make a difference for lunch but does make a difference when bigger purchases are at stake. The only way we’re going to know the answer to “do credit cards make people spend more,” though, is to conduct careful controlled experiments like the ones that these authors have done.