False advertising in LaGuardia

En route to Boston the other weekend, I was hoping to get some work done during my layover in LaGuardia. Rather than finding unlimited free Wi-Fi as I’d been expecting, I discovered that only the first 30 minutes were free. As a cash-strapped college student, I decided I’d do what I could offline and then connect for the 30 minutes if needed. When I finally connected to the free network, I was unable to access anything. I wasn’t entirely surprised, given the amount of traffic that LaGuardia must see on a typical day – 38,000 through Delta’s two terminals alone, according to this estimate.

It later dawned on me that this was an example of private markets acting to more efficiently allocate a public good. While free Wi-Fi is not a pure public good, it is non-excludable and as non-rivalrous as the capability of the server allows. Public Wi-Fi is non-excludable because there is no password and/or price, thus anyone nearby can connect. However, public Wi-Fi is not purely non-rivalrous. The speed of the network will slow down as more people connect, inhibiting the ability of those already connected to “consume” the Wi-Fi. With only one public network, users who are scrolling through Facebook and Instagram constrain the ability of those trying to utilize the network for work. When one considers the individual utilities of such proxy groups, SMC probably exceeds SMB. Thus, by imposing a cost to access Wi-Fi beyond 30 minutes - and an even higher cost if you want extra high-speed internet - the market allocates the good more efficiently. Thanks to this market mechanism, I was able to write a cover letter in record time. I wasn’t about to shell out $10+ knowing that I’d likely end up being the mindless millennial scrolling through her social media feeds.

What I would’ve tweeted @LaGuardia airport had I decided to pay: being able to connect to but not actually utilize a wireless network does not constitute “free Wi-Fi.”

Finding Prime Real Estate in Potomac Mills

            I went to Potomac Mills over the reading days to buy more business professional clothing for upcoming job interviews. For those who do not know, Potomac Mills is one of Northern Virginia’s least illustrious strip malls, but it is unparalleled in cost effectiveness and proximity to my parent’s home. Additionally, all the stores you could hope to find business professional clothing in are grouped together. Ann Taylor is next to Banana Republic, which sit across from Brooks Brothers and J.Crew. The products sold at each store are homogeneous. You can buy a brown, blue, or black suit, with barely a button out of place to differentiate the products, for about $200 at each store. The signs hanging in the window of each retailer advertised 30%-50% off of full priced items, sales attempting to beat out competitor prices between the stores.

            These stores demonstrate how competitors with homogeneous products will group together in order to maximize their demand base. As Hotelling mentioned in The Economic Journal, each retail store moved closer and closer together in order to maximize the amount of customers they reached since they were selling identical products. The stores attempt to choose the location and quantity price that allow it to have access to the most extensive section of the market. Ann Taylor seems to have claimed the prime location to do so, with placement that gives its store the greatest access to the rest of Potomac Mills in relation to the other stores mentioned. The marketing strategy did benefit the store since it was the most accessible women’s clothing retailer in Potomac Mills, and I left the mall with a shopping bag full of Ann Taylor apparel. 

Voting Factors

I, too, decided to go home during reading days, and when I got back to Salem I started to think about voter rationality. Salem is a small independent city in Southwest Virginia with about 25,000 people. As such, the whole community is pretty intimate - I can't go anywhere without seeing someone I know or having some random person text my parents that they saw me out. Salem is right next to Roanoke City, which is a bit bigger with around 100,000 people. Given the population differences and variance in community feel/climate, I started to wonder if the voter turnout rates would be different between the two localities.

In our voter rationality model, we concluded that it was rational to vote if the expected marginal benefit of voting + some variable D, that could represent additional factors such as lessening social pressure and expressive voting, was greater than the marginal cost of voting. My initial hypothesis was that in Salem, the value of D would be higher given the increased social pressure based on the intimacy of the community. Additionally, I figured that expressive voting would be more valuable in Salem than in Roanoke given the importance of social capital in a small town. Therefore, assuming that the expected marginal benefit from voting for the average Salem citizen is roughly similar to the expected marginal benefit from voting for the average Roanoke citizen (economically/geographically similar with same very small p of being deciding vote) and the marginal costs of voting are similar (proximity to polls, time to vote), then I would expect voter turnout to be higher in Salem than in Roanoke as on average the expected marginal benefit from voting + D would be greater in Salem. Voter turnout data for the 2016 presidential election supported my hypothesis as the voter turnout in Salem was 71.55% of registered voters versus 62.7% of registered voters in Roanoke City. However, it is possible that voter turnout in Salem is higher also because of lower marginal costs of voting, meaning it may be the combination of lower marginal costs and higher values of D that explain the difference in turnout.

Boulder's Soda Tax and Foot Voting

I went home to Boulder, CO for fall break for the first time in a while. Boulder is a unique city, and upon arriving back home, I found myself surprised again by the mix of the public goods Boulder provides. One example of this public good is Boulder’s sugary-drink tax, a two cents per ounce excise tax on sodas that voters passed in the 2016 election — the highest soda tax in the country, and one of only eight in existence. Additionally, to protect the mountain views, no buildings are allowed to exceed 55 feet in height.

These laws are unique in nature, and their passage was controversial — I was certainly annoyed by the higher cost of my soda when I got lunch for the first time since getting back. However, it struck me that these are the exact types of laws Charles Tiebout would have advocated. Boulder is providing a unique set of public goods to try and attract residents — allowing them to vote with their feet. Neighboring towns Lafayette, Longmont, and Louisville have neither of the above laws. Individuals should be able to choose where they want to live based on these laws — if people want cheaper sodas or taller buildings, they can live elsewhere. By Tiebout’s standards, these laws appear to be a success, as Boulder is growing at a rate well above the national average. However, Tiebout doesn't account for overflow of public goods to other localities (among other things) in his model, as I can live closer to the mountains in Boulder, but drive ten minutes to Louisville for cheaper sodas. Additionally, people in Louisville benefit from the unblocked views Boulder's building laws create, but can live in taller buildings themselves. Thus, factors not accounted for by Tiebout’s model may affect individuals living decisions, not necessarily Boulder’s unique set of laws themselves.

Sanders-Trump voters illustrate multi-peak preferences on Down's traditional scale

It is estimated that somewhere between 6 and 12% of those who voted for Senator Bernie Sanders in the 2016 Democratic primary ended up voting for Donald Trump over Hillary Clinton in the 2016 general election. That number is large enough in several very close states that had this group voted for Clinton, they would have brought her over the top to win the election. This statistic is very baffling to many as Bernie Sanders certainly seems closer to Clinton in ideology and on the left-right spectrum Downs uses than he is to President Trump, who is on the opposite far end. Using that spectrum, these Sanders-Trump voters would illustrate multi peak preferences, meaning they prefer a candidate seemingly very different from their ideal preferences/candidate to one only slightly different. This violates the single-peaked preference assumption of Downs's model.

However, I propose a different spectrum for understanding this set of voters that would make their preferences single peaked once again and their choice to vote for Sanders and then Trump rational. That is, these voters didn't necessarily care about a candidate's ideological purity, but rather the extent to which they saw them as a change to the status quo in Washington. They preferred the candidate that they viewed as most likely to shake up the system (i.e. draining the swamp, not being beholden to the money of big business, etc). You can argue whether Bernie Sanders ranks higher than Donald Trump in that regard, but Trump undoubtedly ranks higher than Clinton, who essentially ran on the idea of continuing the Obama legacy and has been in government for decades. Thus once Bernie Sanders was out of the running, Trump was the next best choice for these voters who highly valued upending the political system. Those who supported Sanders in regards to their preference for leftist ideology, in turn, still voted for Clinton (or, in some cases, Jill Stein as a protest vote to push Clinton to the left).

The increasing number of and preference for populist, anti-establishment candidates suggests we may benefit from rethinking the traditional left-right spectrum to better understand voter preferences and behavior.

Is it rational to give to political campaigns?

The other day my politics seminar discussed campaign finance. Elections are by definition won by getting the most votes, but too often the reality is that money buys elections. During this seminar I was naturally thinking about Public Choice. We had just covered that voting isn’t rational, assuming that the only utility gained from voting is outcomes driven by policy and that you cast the deciding vote. So I wondered whether donating money to political campaigns, contributing the money needed to buy the election, was rational or irrational in a similar way to voting. I started scribbling a model in the margins of my notes and I reached the following:

|V2 - V1| C/F > C - B

The first notable difference is that the probability of your vote being the deciding vote has been replaced by the probability of one one of your dollars being “the deciding dollar.” C/F is your contribution divided by the total funds the campaign puts towards the election. C-B is the cost of giving, your contribution minus any benefit gained from giving; I had tax write-offs in mind, but this could also presumably be personal altruistic benefit, access to fancy donor cocktail parties, or whatever else.

How rational it is to give to political campaigns depends on how much gain one would receive for one candidate winning over the other and how much other money the campaign has raised. If you’re an owner of a large oil company and one candidate wants to require all new cars to be electric, you will have a massive monetary incentive to help the other candidate to win. Further, if the candidate you prefer doesn’t have a lot of funds, your incentive to give further increases. For the average citizen however, it looks a lot like voting, so I wouldn’t recommend spending your next paycheck on helping your favorite candidate win.

Minimax Regret and my flight home

As we all enjoy these “reading days,” many of us will be leaving C-Ville to travel. Yesterday, I flew home to Connecticut on the 8pm flight. Now anyone who has been through airport security knows it’s a crap-shoot whether the lines are unbearably long or not. My father, ever the pessimist, is a firm believer in arriving 3 hours early to the airport for international flights and 2 hours early for domestic flights, to ensure you make it through security in time. Because of this philosophy, I have spent many family vacations sitting in an airport gate, hours early, waiting for a flight that would likely be delayed anyway. Don’t tell my parents this, but instead of using my dad’s hard and fast rule of “2 hours early,” I decided to use the minimax regret table instead to decide what time I should leave for the airport.  The two alternatives are 1) leave late 2) leave 2 hours early. The two states are 1) make the flight 2) miss the flight.

The cost of my ticket was around $200, but this is a sunk cost. However if I missed my flight because I left late that would be like losing $500,000. This is because my parents would NEVER let me live down the fact that I missed my flight home because I recklessly left my apartment too late. However if I left too early, and was stuck waiting at the gate, this would be pretty inconvenient. By following the “2 hour rule,” I would probably be waiting for one and half hours at the gate minimum. If I value my time at $20 (my usual babysitting hourly fee), then it would cost me around $30. However, going home is worth at least $100,000 because this is one of the last times I can visit before my parents sell the house and move to South Carolina. Therefore, this event would be worth $99,970 in total. If I leave 2 hours early and still miss the flight, I can at least defend myself against my parents and say that I tried to follow my dad’s advice. This would save me the $500,000 cost of my parents scolding me, so there would be no regret. Using the table, it is obvious I would minimize regret by following my dad’s advice and leaving 2 hours early. Sadly this exercise taught me that my parents are always right.

	Make the Flight	Miss the Flight
Leave Late	0, no regret	($500,000)
Leave 2 hours early	$99,970	0, no regret

The Stasi and Rational Ignorance

After Public Choice, I go to a history class called “The Fall of Communism.” The same day we discussed rational ignorance, I learned about international and domestic spy networks during the Cold War. Amidst Cold War fears of “capitalist encirclement,” Soviet republics watched their people constantly. Millions’ daily lives were under surveillance, often by friends and colleagues. Fueled by paranoia, the idea was that in order to maintain a dictatorship, you had to know everything about the lives of others.

Throughout its existence, 1 in 40 Germans informed for the Stasi, the German Democratic Republic’s Ministry for State Security. There are over 100 miles of files saved. Here is an article, published 9 months after the Berlin Wall fell, about the shocking discoveries of the Stasi’s deep infiltration. The Stasi is referred to as the “most sophisticated and far-reaching espionage ever created.” However, my professor argues that the Stasi gathered so much information that it was ultimately difficult to make anything useful of it. The Stasi was omniscient, gaining knowledge to a point where the marginal cost of doing so was far greater than the marginal benefit. The Stasi exceeded their optimal amount of surveillance. Among the many inefficiencies and perversions of communist regimes, these systems were also not rationally ignorant.

A similar analysis could be applied to big data today, though the marginal costs and marginal benefits are certainly different than they were in East Germany. Considering the privacy debates of the 21st century, it is interesting to ask what the optimal amount of domestic intelligence, or rather ignorance, is in the United States. 

The City of Berryville vs Mike of Mario's Pizza

My small hometown has been in chaos this past year over a McDonalds. I am from Berryville Virginia where we pride ourselves on being built and sustained by small local business and agriculture. It was established in the 1700's and the residents value maintaining those historic roots. For about 10 years, McDonalds has really pushed to build a restaurant in Berryville because the town is strategically placed a long a highway that many commuters take to get to Northern Va and DC. The Town Council rejected this proposal over and over, until January 2017. Early this year they voted to allow the McDonalds chain to finally enter into our historic town, despite very loud complaints from the townspeople. After this surprising approval, it was discovered that the land the McDonalds wanted was owned by a local businessman, Mike of Mario's Pizza, and the local government in fact did not have the authority to allow the McDonalds in.

As the link suggests, the townspeople began a petition to both the Town Council and to Mike urging them to refuse the land to McDonalds. The petitioner claims that if Mike does sell, it is because he doesn't care about the fate of his own business or about upholding the character of the town. I think ultimately, Mike will sell the land to McDonalds, not because he wants our town or his business to crumble, but because it could be a rational choice for him. If the benefits of selling his land outweigh the cost of the increased competition and social pressure from the town, it would be rational for Mike to sell, despite the fact that the McDonalds will be a negative externality on the townspeople (although they may change their minds when they get a $1 coffee). Mikes decision will not be based on a like or dislike for his town, but on how each option will effect his personal utility. The town as a whole may be worse off, but private property rights are preventing a more efficient solution. I personally would prefer if the McDonalds did not come into my small town, but I also don't think that it's fair to demonize poor Mike of Mario's Pizza for making a rational decision.