Food Deserts as Market Failure

NPR and others have been reporting for some time about "food deserts," areas that are underserved by grocery stores. These "deserts" are usually, but not always, concentrated in low-income areas. A new story highlights "banking deserts," in which a similar paucity is observed in banking services. Once again, the poor bear a disproportionate share of the burden.

A bog-standard free-market analysis of this problem holds that these firms (grocery stores, and now financial institutions) observe marginal revenue below average variable cost. In other words, the firm identifies locations at which the day-to-day costs of operating the location are greater than the day-to-day revenue generated by the location. The firm then decides that it is profitable to shutter the location. Causal explanations for this phenomenon usually invoke higher shrinkage rates (theft, spoilage), or basic demand deficiencies (the poor would rather eat fast food) for the services involved. Further, market fundamentalists say that residents of the affected areas who desire the abandoned services are incentivized (and therefore likely) to relocate to areas where such services are more readily available.

This analysis fails to account for the often high costs involved with relocation. Let's try a thought experiment. If we disaggregate the population of a low-income food desert into "employed" and "unemployed," we can quickly see the positive correlation between "employed" and "available financial resources to relocate." We may even be able to infer a positive correlation between "employed" and "demand for fresh food," though this controversial and problematic inference is not material in this thought experiment. 

Regardless, let's assume both of these correlations hold for now. If those who most want the fresh food are most able to relocate, then perhaps these closures are the efficient outcome. Stop for a moment, however, and consider our relocation criteria. We observed that the employed have the most resources at their disposal in order to move. On the other hand, the employed have the most to lose by moving, namely, their jobs. I think it's safe to assume that the employed fraction of a low-income population have the most tenuous hold on their jobs of any employed person (who isn't the drummer for Spinal Tap). Often these workers must commute far to their jobs, and often by public transport. Unless the representative low-income employed person who seeks to relocate to escape a food desert has the good fortune of being willing and able to move closer to their job, it's unlikely that any fraction of the population of "food deserts" is made better off by these closures.

Is the best solution to this problem the "government takeover of grocery stores" that Mr Johnson so fears? Probably not. Thankfully, no one is suggesting it is. If, on the other hand, you believe that we elect a government to fix problems, then a small federal program to incentivize grocers seems like a staggeringly pragmatic solution.



See also:

Schuetz, Jenny & Kolko, Jed & Meltzer, Rachel, 2012.

"Are poor neighborhoods “retail deserts”?,"

Regional Science and Urban Economics,

Elsevier, vol. 42(1-2), pages 269-285.

Links of the Week - 15 November 2013

Australia had a government shutdown once. In the end, the queen fired everyone in Parliament
Max Fisher

Modeling Policy When Policy Is Inside the Model
Peter Dorman

World War II in Europe: Every Day

CEOs All At Sea
Paul Krugman

Alan Greenspan Rediscovers Keynes—Sort Of
John Cassidy

Is There Any Appropriate Reaction to This on the Part of Eugene Fama?
via Brad DeLong

What frequent flyer miles teach about central banking
Neil Irwin

The tea party’s assault on workers
Mike Konczal

Adjusting the Taylor Rule for the Unemployment Rate Bias
Jared Bernstein

A Tale of Two Fed Presidents
Paul Krugman

Bubbles of fun

Evan Jenkins, sitting in for Guan Yang, in turn sitting in for Noah Smith, writes, in a long exposition on Bitcoin, that it surely represents a bubble at this time. I want to focus on the two "confounding factors" Jenkins cites. I'm not sure if he meant that these factors confound his analysis, as I don't see that. It's more that these two facts confound the entire Bitcoin project (if "project" is the appropriate term).

Jenkins: (emphasis added)

The first confounding factor is that the Bitcoin market, as it exists now, probably does a pretty bad job of pricing correctly. The problem is that it is very difficult for somebody who does not already hold Bitcoins to make a bet against the future of Bitcoin. The only people with the power to move the Bitcoin market down are those who have already bought into the Bitcoin market, and they likely have rosier visions of the future of Bitcoin than the rest of us. This, in my opinion, is the biggest piece of evidence in favor of calling Bitcoin a bubble. In order for the Bitcoin market to price Bitcoin correctly, there needs to be a good way to short Bitcoins. But until there is a reliable way to lend and borrow in Bitcoins, that won’t happen. 

The second confounding factor is that if Bitcoin really does establish itself as a legitimate currency, we will need to throw our assumptions of how many Bitcoins there are out the window. In principle, there should be no more than 21 million Bitcoins ever produced. In fact, since Bitcoins can be irreversibly lost, we should actually expect the number of extant Bitcoins to start decreasing at some point. But, as much as the True Believers like to rail against fractional reserve banking, the truth is that once Bitcoins become a real currency, it will start being banked and invested like a real currency, which will effectively increase the number of Bitcoins in existence. Depending on how we measure total “amount of money” in US dollars is anywhere from 10 to 20 times the amount of physical currency in circulation. But there is reason to think that for Bitcoin, the mulitplier could be even higher, as Bitbanks would lack, at least initially, any sort of reserve requirement. This lowers the future price, and thus raises the probability, of the True Believer future. But is this the future that the True Believers want? Not really.

To my view, what Jenkins is saying here is that a future in which Bitcoins is a real currency that competes with what we think of now as real currencies, without the bummers inherent in real currencies like fractional reserve banking and governmental oversight (or literal control), is not a real future. In other words, the things that define modern currencies are the things that Bitcoin advocates seek to avoid. This is likely problematic if these advocates want to see Bitcoin compete with real currencies. If, on the other hand, Bitcoin settles into a comfortable role as a niche asset, like tobacco shares or Renoir paintings, then it will likely be a success.

Bitcoin is a neat throwback to the days of state banking with its minimal oversight, competing currencies, high transactions costs, wild price swings, and regular financial panics. Anyone who misses those days of sweet freedom is likely a Bitcoin advocate. None of this is to say that the system we have today is free from frictions and inefficiencies. My point is that, as fuddy-duddy as 21st century currencies like the dollar and the euro are, and as exciting as Bitcoin promises to be, who really wants their medium of exchange, their unit of account, or their store of value to be fun?

Response to Mankiw on community rating (in the unlikely event his questions were genuine)

Greg Mankiw asks "Is Community rating fair?" His contention seems to be that the market has deemed that people with different health risks should pay more for health insurance, much in the way that the market has deemed that people who are statistically more likely to be involved in auto accidents should pay pay more for car insurance and that people who are statistically more likely to die sooner should pay more for life insurance. By way of example, he cites teen drivers (higher car insurance premiums) and old people (higher life insurance premiums, if they can get a policy at all).

Mankiw's complaint is specifically that pregnancy is, for the purposes of the ACA, a "pre-existing condition," the cost of which should not be spread among all of the insured, because pregnancy is a choice. Leaving aside the fact that the degree to which pregnancy is a choice is correlated with, and proportional to, socioeconomic status, Mankiw takes a logical leap that goes something like this:

1. Pregnancy is a choice.
2. Pregnancy is a pre-existing condition.
3. Community rating spreads costs for pre-existing conditions across all insured.
4. Community rating is unfair!!

It should be noted that community rating is primarily meant to remove the penalty of a lifetime of burdensome costs for those of us unlucky enough to be born with genetic conditions, to contract less-than-curable lifetime ailments like Lyme disease, and other catastrophic health scenarios that currently send people down a road toward something that most closely resembles serfdom. It should also be noted that community rating is very similar to the way in which employer-based health insurance plans spread risk, and have done so for decades. It's not like we're landing on Neptune here. Perhaps pregnancy is different in some fundamental way from the sorts of conditions I describe. Then again, maybe it's not. This is a relatively simple technocratic question, not the grand philosophical debate over which Mankiw wastes so much hand-wringing. Perhaps, in a sane political environment, we'd be able to discuss this question and others like it calmly and reach a practical solution.

So, Professor Mankiw, since you asked, yes, fairness is being treated differently here. And further, since you also asked, it's because This Thing isn't like Those Things. And we can't have a rational conversation about this stuff at the legislative level because your political masters can't talk about this stuff without bringing up death panels and the Fugitive Slave Act.


UPDATE: Matt Yglesias adds his inimitable take.

(The Return of) Links of the Week - 8 November 2013

The Derp and Fall of Inflation Fearmongers
Matthew O'Brien

Do the inflationistas really believe what they say?
Noah Smith

Down With Lemonade Stands
Michael Lemberger

Regions of Derpistan
Paul Krugman

Show me the model
Antonio Fatas

Post-Scarcity Economics
Tom Streithorst

The Federal Reserve helped fund the D.C. Metro system. Wait, what?
Ylan Q. Mui

The Case for Paying People More
Justin Fox

Getting paid: Labour markets and partisan minds
Will Wilkinson

This CNBC clip shows everything wrong with how Wall Street thinks about banking
Neil Irwin

FedSim Followup

In my second attempt at Evan Soltas' FedSim, I was able to hold off rising inflation expectations for a while, mostly by raising the Funds rate when the pace of improvement in the labor market was encouraging, rather than when the level of unemployment was satisfactory. The cost of this, of course, is that I stifled that pace, and we found ourselves with unemployment stuck between 6 and 7%, and inflation, especially core, began to pick up significantly toward the end of the experiment. These are, of course, Milton Friedman's "long and variable lags." Oh, well, back to the woodshed.

Showing improvement

FedSim, by Evan Soltas

Blogger, college student, Wonkblog contributor, and all-around wunderkind Evan Soltas has created a monetary policy simulator called FedSim. The idea here is to do a better (more realistic) job of presenting a simulation of central banking action than those that are offered by the ECB and the San Francisco Fed.

In a followup post, Soltas outlined the gargantuan obstacles to the sort of modeling that underlies these sorts of simulators. I think he's all around done a great job overcoming them, however.

My first attempt was going along fine until Zimbabwe set in. I think I allowed the Funds rate to stay so low so long that my 'citizens' came to anticipate low rates forever. Thus my wild punches (2020 and onward) at inflation never landed. My next attempt, the results of which I will also post to CW, will hopefully be less Weimaresque.

I'd encourage readers to try the simulator out and see what you get.

First sad attempt

I am Emmanuel Saez.

I've been relatively neglectful of this blog over the last several weeks. I hope that is set to change, but I make no promises. In the meantime, enjoy this diversion.

In the interest of transparency, here are my answers.

Success and the perception of success

Ezra Klein writes that "Obamacare will never be a huge political success for Democrats," in response to Yglesias and Krugman, who are more optimistic. Klein believes that this is because the benefits to health care reform will be largely invisible to the majority of Americans, and that the costs will be "marginal." I mostly agree, but I will take an even more pessimistic tack by noting that, for the foreseeable future, some people will blame any poor healthcare outcome on the new law. Price of an MRI goes up? Obamacare! Longer wait for appointments? Obamacare! Unfortunate test results? Obamacare! Weird guy gave me the side-eye on the way to the doctor's office? Obamacare!

Now, it is likely that this group will never comprise more than five or ten percent of the population. Remember, however, that this group is likely only that big now, and yet they control two of the three branches of government.

Links of the Week - 28 June 2013

Short list again this week. -ed.


Robert Solow:
Hayek, Friedman, and the Illusions of Conservative Economics

Jared Bernstein:
The Top 1% and the College Earnings Premium: Don’t Confuse Them (as Mankiw Does)

Jared Bernstein:
Inequality and Mobility, Again

Noah Smith:
Some essential papers in behavioral finance

Links of the Week - 21 June 2013

Kevin Drum:
Chart of the Day: America's 30-Year Project to Make the Rich Even Richer

Mark Thoma:
7 Important Examples of How Markets Can Fail

Paul Krugman:
Rents and Returns: A Sketch of a Model (Very Wonkish)

Jonathan Chait:
Why Rand Paul Distrusts Democracy

Jonathan Chait:
Is Rand Paul’s Love of Ayn Rand a ‘Conspiracy’?

Harold Pollack:
The complacency of the meritocrats

Paul Krugman:
Greg Mankiw and the Gatsby Curve

Elias Isquith:
Rand Paul: Not Aristotle

The Economist:
The 1 percent needs better defenders

Links of the Week - 14 June 2013

A light list this week. -ed.

Jared Bernstein:
No, Public Sector Jobs Do Not Crowd Out Private Sector Ones

Neil Irwin:
How the Spurs beat the Heat: the principle of financial arbitrage

Paul Krugman:
Unproductive Finance

(related: Karl Smith: How Schlubs Get Taken By Wall Street Pros)

Neil Irwin:
Alan Krueger on how the music industry explains inequality

Jared Bernstein:
Pictures of the Corresive Linkage Between Higher Inequality and Diminished Opportunity