Wednesday, October 31, 2018

Why are prices going up?

WSJ reports that prices are going up because costs are increasing and income is increasing, increasing comsumers' willingness to pay:

Businesses including Coke and big U.S. airlines have said their higher prices aren’t denting demand
“The economy is healthy,” Delta Air Lines Inc. Chief Executive Ed Bastian said in September. “To the extent oil prices were to continue to rise, we expect to be able to pass along the cost of that.”

Monday, October 29, 2018

Platforms as Market-makers

There are two different ways of looking at markets: metaphorically and as "platforms."  Metaphorically, markets are ways of characterizing transactions between a group of buyers and sellers for a given product, in a geographic location, and during a specific time.  There is no literal marketplace, but prices and quantities behave as if there were.  

Platforms are literal markets that are under the control of a market maker, like Uber, who can set different prices to buyers (riders) and sellers (drivers).  For example, suppose that there were seven sellers with marginal costs {$1,$2,$3,$4,$5,$6,$7} and seven buyers with values={$7,$6,$5,$4,$3,$2,$1}.  Uber makes money on the spread between what it receives from riders and what it pays to drivers.  With these supply and demand curves, we show Uber
s profit calculus below.


·       If Uber paid $1 and sold at $7, it would sell one ride for a profit of $6. 
·       If Uber paid $2 and sold at $6, it would sell two rides for a profit of $8. 
·       If Uber paid $3 and sold at $5, it would sell three rides for a profit of $6. 
·       If Uber paid $4 and sold at $4, it would sell four rides for a profit of $0. 

The profit maximizing prices are in red above.  The table also shows how the competitive equilibrium (a single price) is actually a special case of the platform with zero margin. 

This raises an interesting question, is it ever profitable to bring an extra driver into the market?  In this simple example, the answer is no.

However, the answer changes if bringing more drivers into the market increases the riders value for rides by reducing the waiting time.    For example, suppose that an extra unmatched driver into the market reduces waiting time by enough so that it increases the willingness to pay by every rider by $2, i.e., with values={$9,$8,$7,$6,$5,$4,$3}.  This would change the profit calculus as follows:


·       If Uber paid $2 and sold at $9, it would sell one ride for a profit of $7. 
·       If Uber paid $3 and sold at $8, it would sell two rides for a profit of $10. 
·       If Uber paid $4 and sold at $7, it would sell three rides for a profit of $9. 
·       If Uber paid $5 and sold at $6, it would sell four rides for a profit of $4. 

With an extra unmatched driver, the margin that Uber can charge increases by $2, which is enough to offset the cost of bringing an extra driver into the market.  In fact, these kind of network effects work in both directions.  Just as bringing more drivers into the market reduces the waiting time for riders (increasing demand), so too does bringing more riders onto the platform make it easier for drivers to find nearby riders (increasing supply). 

What is the elasticity of demand facing Whole Foods?

In the FTC's 2006 suit to block the merger of Whole Foods and Wild Oats, former colleague Dave Scheffman argued that the simple pricing model of Chapter 6 implied that the merged firm would not find it profitable to raise price:
In his decision, the Judge cited a break-even analysis (sometimes called "critical loss") that asks how much quantity a monopolist could afford to lose and still want to raise price. With retail margins of 40% (FTC complaint, p.21), a 5% price increase would require a quantity loss of no more than 11.1% to be profitable [11.1%=5%/(5%+40%)]. Citing marketing studies showing that customers shopped at Whole Foods as well as other grocery stores, former FTC Chief Economist and colleague David Scheffman argued that the actual quantity lost would be greater than 11.1%, presumably to stores outside the category.

Implicitly, Professor Scheffman was using the (P-MC)/P=1/|elasticity| to infer an elasticity of -2.5.  We can compare this elasticity to the the one we computed from the pricing "experiment" after Amazon purchased Whole Foods for $13B.  Amazon's pricing algorithms (presumably they used algorithms to compute prices) reduced prices by 30% and foot traffic increased by only 3%.  If foot traffic is a proxy for sales, then we can calculate a price elasticity:

 elasticity = (%change in Q)/(%change in P)=(+3%)/(-30%) = -0.1

Either this elasticity estimate is way off (foot traffic is a bad proxy for sales), or Amazon is trying to accomplish something other than the short-run profit maximization of chapter 6. 

Friday, October 26, 2018

Japanese towns are "bidding" for citizens

Japan's productive cities like Tokyo are growing, and the smaller towns and cities are shrinking:
Educating children is incredibly expensive. The regions are quite annoyed that they pay to educate their children but that [cities like] Tokyo reaps all the benefits. This state of affairs has continued for decades.

A substantial part of Japan's income taxes (8%) are paid to the city that is designated as the citizen's "hometown."  The program is called "Furusato Nouzei" and
... allows you to donate up to 40% of next year’s residence tax to one or many cities/prefectures of your choice, in return for a 1:1 credit on your tax next year. This is entirely opt-in. Anyone can participate, regardless of where they live.

When the program was created, it appeared that it wouldn't be very popular because people have to take affirmative steps to transfer the money, and they have to float the money for a year before they get a tax credit in the following year.

But this program created an incentive for cities to "bid" for taxpayers to select them as their "hometown."  Internet sites popped up, and rewarded taxpayers with "travel vouchers" that were almost equivalent to cash.
...many rural towns without large expatriate (inpatriate?) populations and without much to differentiate them in terms of local food had bid the consideration for a donation up, up, up. 

Eventually, the central government had to step in to "cap" the amount that a city could bid for a taxpayer:
...the maximum they’d allow is you rewarding a taxpayer with 50% of the donation in consideration. So, if you “donate” ~3% of your gross salary to one of these cities (which is 1:1 matched by e.g. Tokyo; you’re donating someone else’s money), they will give you ~1.5% of it back in all-but-cash.

I wonder how high the bids would have gone (would all of the tax been returned to the citizens?) if the government hadn't stepped in?  I am going to blog this under "bid rigging" which, which if done by private parties is a criminal violation of the antitrust laws. 

Thursday, October 25, 2018

The Opportunity Cost of Socialism

New report by the Council of Economic Advisors on the effects of socialist policies.  Here are some highlights:

In assessing the effects of socialist policies, it is important to recognize that they provide little material incentive for production and innovation and, by distributing goods and services for “free,” prevent prices from revealing economically important information about costs and consumer needs and wants. To this end, as the then–prime minister of the United Kingdom, Margaret Thatcher (1976), once argued, “Socialist governments . . . always run out of other people’s money,” and thus the way to prosperity is for the state to give “the people more choice to spend their own money in their own way.”
...This literature finds a strong association between greater economic freedom and better economic performance. 
  • It suggests that replacing U.S. policies with highly socialist policies, such as Venezuela’s, would reduce real GDP at least 40 percent in the long run, or about $24,000 per year for the average person.  
Although they are sometimes cited as more relevant socialist success stories, the experiences of the Nordic countries also support the conclusion that socialism reduces living standards. In many respects, the Nordic countries’ policies now differ significantly from what economists have in mind when they think of socialism. For instance, they do not provide healthcare for “free”; Nordic healthcare financing includes substantial cost sharing. ...  Nordic taxation overall is surprisingly less progressive than U.S. taxes. The Nordic countries also tax capital income less and regulate product markets less than the United States does.... Living standards in the Nordic countries are at least 15 percent lower than in the United States.  
  • It may well be that American socialists are envisioning moving our policies to align with those of the Nordic countries in the 1970s, when their policies were more in line with economists’ traditional definition of socialism. We estimate that if the United States were to adopt these policies, its real GDP would decline by at least 19 percent in the long run, or about $11,000 per year for the average person. 

Marginal analysis and choice of majors

MarginalRevolution documents the comparative--not absolute--advantage of boys in STEM subjects:
Loosely speaking the situation will be something like this: females will say I got As in history and English and B’s in Science and Math, therefore, I should follow my strengthens and specialize in drawing on the same skills as history and English. Boys will say I got B’s in Science and Math and C’s in history and English, therefore, I should follow my strengths and do something involving Science and Math.

This analysis done by females can be understood using Marginal analysis, i.e., if MCMath > MCEnglish for females, then the MC of doing math is greater than that of English, so they devote more effort to English.  However, this ignores the hidden benefits of studying Math.  It is more likely that (MRMath-MCMath) > (MREnglish-MCEnglish).  or that the marginal profitability of Math is greater than that of English. The expressions in parentheses are the marginal profitability of each activity.

Wednesday, October 24, 2018

Marginal Analysis of Medicare

We have blogged about over-consumption of medical care by seniors before, as it is a perfect illustration of the marginal analysis of Chapter 4:  if you reduce the MC of an activity, you get more of it.
  • How many times do we have to repeat this until someone listens?
    • taxes for every taxpayer would have to rise by roughly 81% to pay all of the benefits promised by these programs under current law over and above the payroll tax
  • Election update: fiscal wake-up questions for candidates
    • Medicare poses a much bigger challenge than Social Security. According to the 2007 trustees report, Medicare's projected costs over the next 75 years exceed its earmarked taxes and premiums by $34 trillion in present value. Measured as a share of the economy, Medicare will more than double by 2035.
  • Them Geezers is Bleeding us Dry
    • Sure enough, the number of enrollees increased somewhat, but expenditures per enrollee per year increased from ~$1,000 in 1980 to ~$10,000 in 2009. Folks, the automatic increases at this pace just aren't sustainable. I sure wish we could have an honest debate about it.
  • Happy New Year: start worrying now
    • ... an average-wage, two-earner couple together earning $89,000 a year. Upon retiring in 2011, they would have paid $114,000 in Medicare payroll taxes during their careers. But they can expect to receive medical services -- from prescriptions to hospital care -- worth $355,000, or about three times what they put in.
And every MBA student at Vanderbilt has had to watch Stossel's video, on GREEDY SENIORS:

What happened when Amazon ran price experiments to estimate MR?

In 2000, Amazon ran some price experiments to figure out what price to charge.

Consumers were furious:
 A group of consumers caught the online superstore tampering with price tags on DVDs, selling them to different people for different prices. 
Amazon officials immediately went on the defensive, describing the changes as a marketing test and issuing refunds to people who paid more than others. Many customers reacted angrily, wondering whether to believe the company's claim that prices weren't adjusted based on the personal preferences they revealed while shopping the site.

And Amazon learned its lesson:
"We've never tested and we never will test prices based on customer demographics," company founder Jeff Bezos told The Associated Press. Instead, he said, the company was merely trying to find out how much it could charge all customers for certain DVDs while maintaining a respectable volume.

Tuesday, October 23, 2018

Motivating medical device salespeople

Heart valves are sold by putting inventory at hospitals.  When a surgeon opens up a patient, she chooses a valve size based on the patients heart, and replaces the real valve with an artificial one.  When this happens, the sale takes place, and the heart valve company gets paid.

The problem with a particular heart valve company was that its inventory was too big, i.e., it had inventory in hospitals with small expected sales.  Lets run this through our problem solving framework:

1. Who was putting inventory in unprofitable hospitals? 
Medical device salespeople were placing inventory at hospitals with low expected sales. 
2. Did they have enough information to make profitable decisions? 
Yes, they knew the patient volume and the preferences of the surgeons.  
3. Did they have the incentive to do so? 
No, they were evaluated based on sales and paid a 6% commission on revenue. 

OK, now that we understand the problem, lets look at some solutions: 

1. Split the decision rights, so that salespeople would initiate the decisions, but a marketing team would have to ratify the decision, .e.g, based on expected demand. 
This solution ignores the specific information of salespeople, who may know more than the marketing team.   
2. Change the information flow 
It appears that information is not the problem, so not sure this would do anything.   
3. Evaluate sales people based on net revenue, revenue minus the opportunity cost of capital times the amount of inventory necessary to generate sales, e.g., 12% * $1M=$120,000 if the firm's cost of capital is 12% and it takes $1M worth of inventory (the valve company needs to carry a complete set of sizes) to make sales at a hospital.  
This looks like the best solution.   

If we view the hospital as a unit that this can be viewed as an extent decision (at how many hospitals should we place inventory?) for which marginal analysis is used.  The marginal benefit of placing inventory at a hospital is the contribution margin times the expected sales, (P-MC)*Q and the fixed costs of making the sales are F=$120,000.  If the contribution margin is $30,000, then if expected sales are bigger than the break-even quantity, Q=F/(P-MC)=4, then it is worthwhile to place inventory at the hospital.  

Monday, October 22, 2018

Bank stock drops 28%

After $45M worth of loans in its portfolio went bad:

Banks make money by borrowing from depositors (at 1%) and lending to business (at 5%).  A bank's net revenue is the amount of loans they make times the interest rate spread (LOANS*4%).

Typically loan officers have incentives to make as many loans as possible, ignoring the potential for default.  Although banks must raise equity to cushion against the risk of loan default, it is only 10.5% of the value of its loans.  When loans go bad, this erodes some of the owners' capital, which can cause a sharp drop in the bank's stock price.  Fixing this is hard because loan performance may not be realized until years after the loan is made. 

Nate Tobik, founder of, told Benzinga’s PreMarket Prep hosts Friday that Bank of Ozarks has been his favorite short idea for about a year now.  
“I think it’s going to be a zero once we hit a recession, and you can ride it all the way down,” he said. 


Sunday, October 21, 2018

A policy that achieves exactly the opposite of its original goal

Landlords reacted to rent control in San Francisco with three strategies:

  1.  landlords moved into the property themselves, known as move-in eviction;
  2.  landlords evicted tenants to convert the units to condos; and
  3.  landlords offered tenants monetary compensation for leaving.
So who benefitted and who was harmed? control operated as a transfer between the future renters of San Francisco (who would pay these higher rents due to lower supply) to the renters living in San Francisco in 1994 (who benefited directly from lower rents). 

Ironically, the conversion of rental properties into higher-end, owner occupied housing attracted higher-income residents which actually contributed to the gentrification of San Francicso, "the exact opposite of the policy's goal."


Wednesday, October 17, 2018

How much equity should a bank have?

Banks make money by borrowing from depositors, e.g., at 1%, and lending to businesses, e.g., at 4%.  In this case, they would make 3% on the total amount of loans they made.

The above chart shows the increasing levels of equity that a bank is required by regulators carry to as a percentage of Asssets (loans) that it makes.  For example, if a bank makes $100M in loans, it must carry at least $10.5M in equity plus "capital buffers" (last line) to guard against its loans going bad.  With 10.5% buffer, if 10.5% of its loans default, the bank will still have enough to pay back depositors.

You can see why banks want to keep this number low, as it allows them to make more loans, and earn more money.

FASB, the accounting regulators are also trying to make banks realize losses as soon as they know about them, a provision that would reduce their ability to lend as loans default:

The rule, which the FASB issued in 2016, represents a fundamental shift in how banks treat loan losses. Currently, they don’t record losses until they have evidence the losses will actually occur. Supporters of the change say that approach led banks to record losses too slowly after the 2008 financial crisis, leaving investors in the dark. 
The new, upfront method—known as “CECL,” short for “current expected credit losses” and pronounced “Cecil”—is intended to make banks more transparent. But the new method also could require some banks to significantly boost their loan-loss reserves, which could cut into their earnings and regulatory capital.

Some banks say the rule will make them less able to make loans, which in a recession could make bad economic conditions worse. That’s because a bad economy would further boost projections for future loan losses, they say, causing the banks to further increase reserves and leaving them with less capital to lend.

Managerial Econ: What my daughter is learning in Rome

Managerial Econ: What my daughter is learning in Rome: We have blogged about the labor market problems in Italy before...

Sunday, October 14, 2018

Dawn raids uncover evidence of beer price fixing in India

Absent cooperation from one of the conspirators, collusion is difficult to prove because it requires evidence of an agreement between competitors.  Most executives are counseled to imagine what would happen if their correspondence and e mails were made public.  

To gather evidence, the competition authorities create a "prisoners' dilemma", by offering leniency to any cartel member willing to provide evidence against his or her co-conspirators.  In India, it appears that the leniency program lead to evidence of price fixing:
The CCI was tipped off by one of the three companies after it filed a leniency application with the regulator, revealing details of the alleged price fixing, he added. 
The regulator’s leniency program is a type of whistleblower protection offered to cartel members. 
The government source said that the raids found email exchanges showing that the companies were fixing prices. “That is smoking gun evidence,” the source said.
HT:  Matt B.

Thursday, October 11, 2018

Screening on criminal background and credit history

When I was at the Bureau of Economics at the FTC, we were asked by Congress whether using credit histories to price car insurance was discriminatory.  The resulting FACTA report found that:
  1. as a group, African-Americans and Hispanics tend to have lower scores than non-Hispanic whites and Asians.
  2. ...scores effectively predict risk of claims within racial and ethnic groups.
  3. The Commission could not develop an alternative scoring model that would continue to predict risk effectively, yet decrease the differences in scores among racial and ethnic groups.
As a result, banning the use of credit scores would result in insurers finding other, less good and possibly discriminatory methods of distinguishing high from low risks, like selling insurance only in low risk areas.  Good drivers living in higher risk areas would be "pooled" with other drivers living in the high risk area, and would have to pay higher rates.

 Previous studies (here and here) finds an analogous effect of preventing criminal background checks in employment, that doing so increases racial discrimination against African American men:
If employers are very averse to hiring ex-cons then they will seek to reduce this risk and one way of doing so is by not hiring any black men. As a result, a background check allows non ex-cons to distinguish themselves from the pack and to be hired. Furthermore, when background checks exist, non ex-cons know that they will not face statistical discrimination and thus have an increased incentive to invest in skills.
When the box [a criminal back ground check] is banned it’s no longer possible to cheaply level the playing field so more employers begin to statistically discriminate by offering fewer callbacks to blacks. As a result, banning the box may benefit black men with criminal records but it comes at the expense of black men without records who, when the box is banned, no longer have an easy way of signaling that they don’t have a criminal record. Sadly, a policy that was intended to raise the employment prospects of black men ends up having the biggest positive effect on white men with a criminal record.

See also Using credit history to price hospital care