Scott Sumner  

A penny for your thoughts

PRINT
Dianne Feinstein's Misundersta... Reply to Noah on Sheepskin Eff...

Tyler Cowen links to a recent study of Canadian grocery store pricing (the first three lines are Tyler):

There is the danger that prices will be rounded up more than down. A 19-year-old Canadian, Christina Cheung, has done a study of penny abolition in Canada and here are her findings:
She found 60.8 per cent of grocery store prices ended in 8 or 9.

Cheung put those numbers through a computer simulation program that generated hypothetical purchases.

"I simulated [grocery purchases] depending on what province you're in, what tax rate you have, and how many items you buy," she said.

She excluded the volume of credit and debit card transactions, which aren't affected by penny rounding.

She then made a discovery: a typical grocery store would earn an additional $157 of revenue from rounding each year.


I'm not quite sure I understand what is being studied here. At first I assumed that the study had showed that grocery stores usually rounded up after the penny was abolished. On a second look it appears that she predicted this would happen. If so, I'm not quite sure why.

If raising prices would actually raise revenue for grocery stores, then why not do it even before the penny is eliminated? One answer is that it's a game theory problem; this action only benefits stores if all firms raise prices from $9.99 to $10.00. Any firm that did not cooperate would steal business from other firms. But in that case, if the penny is removed then a non-cooperative firm could just lower the price to $9.95 and steal business away from other firms. Sure, they'd lose 4 cents by doing that, but I notice that an increasing number of stores already just give me a few pennies for free to avoid the hassle of counting out each coin. If it's truly a good strategy, then why not price at $XX.95? Wouldn't the failure to do so cost sales; dare I say "penny wise, pound foolish?"

Maybe I'm missing something, but I don't see how this increases the revenue of grocery stores. There would have to be more than 4 increases from $9.99 to $10.00 for each cut from $9.99 to $9.95 in order for stores to get more revenue. Add in the fact that the grocery industry is pretty competitive, which means that if profits increased then the entry of new firms would also increase.

I strongly favor getting rid of the penny, and maybe even the nickel. In 1973, the smallest coin was a penny, which was worth the same as a nickel is today. We got along just fine without a smaller coin. In 1913 the smallest coin was a penny, which was worth as much as a quarter is today. Why do we need small change filling up our pockets, and imposing heavy costs on the government?

PS. In the early years of the country there was a half penny, but it was killed off in 1857 when it had roughly the purchasing power of a dime today. Not needed.

Screen Shot 2017-12-18 at 7.36.35 PM.png
PS. I feel like a horrible person criticizing a paper by a 19-year old student that got published in a scholarly economics journal. (Perhaps I am a horrible person.) But I'd like to think that my critique is a way of showing respect. A penny for your thoughts?


Comments and Sharing


CATEGORIES: Money




COMMENTS (16 to date)
Stefan Krieger writes:

Also fun: the same Currency Act of 1857 that abolished the half cent also banned foreign coin as legal tender. It appears UK and Spanish money circulated widely before that and was legal tender in at least some states.
The Changing Role of Foreign Money in the United States, 1782-1857

Trevor H writes:

If a typical grocery store does a couple million dollars per year in business then an extra $157 is a, uh, rounding error.

As for the method, from that summary my impression is that prices were simply naively rounded to the nearest nickel with no modeling of potential changes in consumer or competitor behavior. Deeper modeling could be done, but if the effect is already trivial I'm not sure it's worth it.

Hazel Meade writes:

I'm guessing the scholarly journal felt equally bad rejecting a paper by a 19 year old student, which is the only reason it got published.

Also, it's worth noting that her assumption is that consumers only buy one item at a time. If you buy three items priced at $0.99 it adds up to $2.97 and gets rounded down to $2.95.

mb writes:

you are missing the results of the rounding on multiples of items, and remember it is rounded to the nearest 5. A store will make money if rounding to nearest 5 increases the charged price. With prices that end in 9, the charged price is greater than the listed price 5 out of 11 times (store makes more money) and equal to the listed price 2 out of 11 times - overall the store only rounds down 4 times. With a price ending in 8, the charged price is greater than the listed price 5 out 12 times and equal to the listed price 2 out of 12 times - overall this is a wash as the store rounds down 5 times as well. With a price ending in 4, the charged price is greater 10 times, equal to 5 times and lower 10 times (similar to a price ending in 8). My guess for the use of prices ending in 8 and 9 is they crunched numbers on typical purchases and found a pricing strategy to maximize rounding up.

Indy writes:

Don't stop at the penny. Get rid of pennies, nickels, and quarters too. And half-dollars and dollar coins. Round all prices to the nearest ten cents. Dimes: one coin to rule them all.

RohanV writes:

I think you're misunderstanding how the rounding rules work in Canada. They're set by the government:

If the customer pays in cash, the final post-tax total is rounded to the nearest 5. If the customer pays via credit or debit, the final post-tax total is not rounded.

I would take the study as more of "did the stores make or lose money due to this change?", rather than "are stores attempting to take advantage of this change?"

Though, if I was looking for stores seeking an advantage, I'd look at coffee-shop prices. You're more likely to have people buying a single item with cash. Most people usually pick up multiple items at a grocery store.

Gwen T writes:

In China, the prices are usually rounded to the nearest renminbi, which is worth 15 cents.

Michael Albert writes:

While we're dreaming, perhaps the money saved in not producing small denomination coins could be used to print bills of differing denomination in different colours?

As the spelling of the last word in the previous sentence indicates, I'm casting aspersions from afar -- here in New Zealand the smallest denomination coin is 10 cents (and has been since 2006, with one and two cent coins having been removed in 1990).

Derrick M writes:

I think the example he used is leading to a conclusion. If a good is priced at $9.99,there is really only one way to go with to get the effect of not using a rounded price. $9.95 is clearly different than $10.05 if this psychological effect is pursues. Alternately, if a good is priced at $1.69, moving to $1.75 might seem more reasonable.

john hare writes:

I get the impression that her point was that evil businesses will use the rounding to exploit customers. A lot of the math in comments seems to miss that businesses will just continue to operate a business without penny gaming. Or IOW, same game, different ante.

ColoComment writes:

Rather off topic, but I once read a mystery novel where an employee of a mutual fund management company conceived a plan to siphon off the final two? digits' worth of value of the 6-decimal price of mutual fund shares into an account that he had set up.
Over time those bits of pennies added up.
I thought it was quite the clever scheme.... ;-)

Ghost writes:

@ColoComment:

See also Richard Pryor in Superman III... ;)

Kenneth Hall writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Justin D writes:

I mostly agree with Indy, though I'd go with dimes and half dollar coins, and drop a decimal place so all prices would be denominated like $7.9 or $14.3. A dime today is worth the same as a penny in late 1950.

I'd also probably get rid of higher denomination bills, the $100 at the very least.

James Pass writes:

As with any model, a lot depends on the assumptions being made for the inputs as well as the quality of the empirical data.

What caught my eye was the line, "She then made a discovery . . ." It's important to know what Christina Cheung was setting out to do. Mr. Cowen states that he isn't sure what was being studied. I'd be shocked to find any research paper published in a scholarly journal that didn't state what was being studied.

The title of the paper seems straightforward: "Eliminating the Penny in Canada: An Economic Analysis of Penny-rounding on Grocery Items." But I can think of various directions such a study could go. Oddly, the paper doesn't seem to have an abstract. I don't have the time to look at the paper in detail, so I can't offer any further comment.

Ossi Saresoja writes:

@Scott:

I'm not quite sure I understand what is being studied here. At first I assumed that the study had showed that grocery stores usually rounded up after the penny was abolished. On a second look it appears that she predicted this would happen. If so, I'm not quite sure why.

If raising prices would actually raise revenue for grocery stores, then why not do it even before the penny is eliminated?

I think the assumption is that the stores won't raise or lower any individual price. However, the total value of your purchases is rounded to the nearest 5 cents when paying with cash (whereas when paying with a card there's no rounding).


@Hazel Meade

Also, it's worth noting that her assumption is that consumers only buy one item at a time.
No it isn't. Although individual prices are mentioned first, it's also mentioned:
"I simulated [grocery purchases] depending on what province you're in, what tax rate you have, and how many items you buy," she said.
Note the part: "I simulated [...] how many items you buy".
POST A COMMENT




Return to top