What killed the local newspaper? Craigslist. The actual reporting function of the newspaper was never what made it profitable -- it was simply a way to get people to look at the classified advertisements. With that gone, the work with positive externalities died. 1/ This is a classic staggered roll-out paper -- if you are willing to buy the assumption that Craigslist expanded to different markets in a way that is unrelated to future trends, conditional on basic demographic controls, then you buy the paper.

Why we so often behave as though the world is zero-sum when it isn't? This paper shows that even the slightest bit of asymmetric information is sufficient to cause us to go away from the optimal outcome, and lead to people to vote for things they don't want in expectation! 1/ Consider three voters, Alice, Bob and Carol. They are deciding whether to enact policy P, and policy P*. P is a payoff of zero to all people. Policy P* gives two voters a pay off of 2, and one voter a payoff of -3. Which voter loses depends on the state of the world.

The marginal propensity to consume (or MPC) is one of the most important parameters to know if we want to estimate the effect of fiscal stimulus. Yet, we may not be able to know the real effect -- even from perfectly identified studies! Stimulus is less effective than we think.1/ Previous micro work on the MPCs take two basic paths. With Parker et al (2013), the identifying variation is coming from random variation in which people actually received their checks.

This paper is an absolutely monumental piece of work. Given incredibly detailed data on how people travel in Chicago, can one calculate the optimal prices of vehicles, roads, and public transport? It turns out -- free buses are correct! And we need massive taxes on vehicles. 1/ To simplify the analysis, we will skip over endogenous reallocation of people, and we are only dealing with pricing the modes which we have now. People are taken to flow between each of the 77 Chicago neighborhoods -- we are not getting more granular than this.

Replacing hospital managers who are doctors with MBAs reduced mortality by 8%. That works out to two fewer deaths per every thousand patients. Managerial quality really does matter, and their pay is justified by the value created. 1/ In 2003, Chile began selecting executives for public programs through an open and competitive process. This would go on to be used to fill 3,400 roles.

Uber famously changes prices in response to demand, hiking prices when more people want the service. Who benefits? The consumer. Drivers lose out slightly compared to a world where Uber sets a uniform price. Total surplus goes up. We shouldn't be afraid of higher prices! 1/ It is not ex ante obvious whether surge pricing will be better or worse for the consumer, or even overall. One can imagine a world where the demand elasticity is changing over time, and so dynamic pricing let's the company charge bigger total markups.

A serious problem in the developing world is that the moment anyone makes anything, their family and relations show up demanding a claim. The "social tax" is so large that 60% of workers take a bank account they can't use for months, just to keep others from knowing! 1/ The experiment is at a cashew processing plant, where workers scrape the plant husk off the nut. The compensation scheme is a pure piece-rate system -- workers are paid by the kilogram. These are almost entirely women.

There's a reason Bukele is so celebrated in El Salvador -- the gangs, MS-13 and Barrio-18, were so awful that they reduced the wages of the people who lived in their zone by more than 50%! People were too afraid of being shot to leave their zone for better jobs. 1/ The El Salvadorean gang situation was an accidental creation of the United States -- when we dumped hundreds of gang members hardened by the streets of 90s Los Angeles, El Salvador could not keep up.

This is *the best* study we have on the effect of phones on academic performance, and an absolute must read. It's really bad for you! It reduces grades and lowers wages coming out of school, with doubling app usage reducing wages by 1.5%, or $135 a year. 1/ It's historically been difficult to measure the effect of videogame usage because it's endogenous. Someone who is already less attentive might play videogames in class -- but who is to say whether it is the videogames causing it, or the inattention which caused both?

The basis of Silvio Berlusconi's political power was his TV network Mediaset. Using the staggered roll-out of the network, the authors find that watching television made young people stupider -- and thus, more likely to vote for Berlusconi. 1/ The fundamental argument behind the paper is that the places where transmitters were placed, conditional upon controls like population and income, were unrelated to future propensity to vote for him.

Listening to the Dave Ramsey Show actually makes people consume less and save more. Using the rollout of the show to different media markets over time, he shows that exposure to the radio show reduces household expenditures by 5.4%. (!!) 1/ If you are not already familiar, the Dave Ramsey Show airs in the afternoon, and is relentlessly focused on personal finances and avoiding debt for consumer purchases.

Contrary to some earlier work, taxi drivers behave like any other worker. They do not engage in "income targeting", where they choose an amount they want to make in a day and work till that is reached, which would imply a negative response of hours worked to income shocks. 1/ Back in 1997, a group of leading behavioral economists thought they had found that the elasticity of labor supply was negative, which would track best with workers choosing a target. If wages are unexpectedly high, then they will work less.

In a recent op-ed, Prof. Jean Twenge claims that college students who took notes by hand had way better academic results. This would be incredible -- if it weren't for the fact that that's not what the study shows! This is a serious act of misrepresentation. 1/ The meta-analysis she cites is quite annoyingly paywalled, but I did eventually get to read it. The studies they are analyzing are standard lab experiments on undergrads for notetaking. They have two groups, and have them answer a quiz after watching a 15 minute talk.

Especially in the developing world, technological advancements are not used for puzzling reasons. When a group of Ivy League economists invented a better way to manufacture soccer balls, they couldn't get anyone to take it. Sialkot, Pakistan, is one of the foremost centers of soccer ball production in the world, and it all started with a chance encounter between a British soldier and a saddlemaker. By 2012, 135 firms employed around 36,000 people to produce 20% of the world's soccer balls.

In one of the more jaw-dropping papers of the year, these economists used the complete consumption data of every single person in Norway from 2006-2018 to answer the biggest question in contemporary macroeconomics: how much does inequality matter? Maybe -- not so much! 1/ So how did we get here? The standard New Keynesian approach was to assume a representative agent, with the average of each population characteristics. Finding the aggregate impacts of policy was easy -- just solve for the household, and multiply by the number of people.

How should we conduct monetary policy? Remember that, if firms could costlessly adjust, monetary policy would be irrelevant. Therefore, optimal monetary policy depends on the precise way in which firms are unable to adjust. This paper is one of the key pieces of evidence. 1/ We are going to have to detour a bit to explain the context. As noted, if prices could perfectly adjust, then monetary policy doesn't do anything. To fix this, we add in some form of stickiness. There are a few basic ways to do this.

Logarithms are ubiquitous in economics. They allow you to conveniently express percentage changes. However, your data includes a zero, the log representation breaks. It turns out that there is no solution fully robust to the choice of units used! 1/ Suppose you’re testing the effect of pay on hours worked. You have random assignment of the program, so you’re not concerned about identification. You find that it increases hours by 2 hours per person on average — no problem, the average treatment effect is easy to represent.

Myerson-Satterthwaite proved that efficiency through trading will never be reached, even if there are no costs to engaging in a transaction. The way in which goods are allocated initially really matters, and thus, so does auction design. 1/ Suppose two parties, A and B, are to trade a good. There are a few desideratum we would like to have — we’d like participating to be individually rational, we’d like to have the person who values the good more have it, we’d like no subsidies, and we’d like incentive compatibility

Total factor productivity growth is slowing down, but we have more patents than ever. What gives? Aakash Kalyani argues that if you look at the actual text of the patents, they’re getting less creative — a change likely caused by falling population growth. 1/ How can you tell creativity? Obviously, we cannot have RAs read all the patents, but we can use simple computational methods. Find all two word combinations, and discard those in common use before 1900. Those are technical bigrams. The less they’ve been used, the more creative.

This is one of the most extraordinary papers ever written. Economists have long wondered how much the growth of cities is due to spillovers. The authors use the fall of the Berlin Wall to show that while spillovers are substantial, they’re very local. Let’s understand how 1/ First, the question. There are two basic reasons why people might be where they are. It could be that the location is really good for a city, or it could be that there are already people there.

This paper is one of the most astonishing feats of sustained data wizardry I have ever seen. Using data from Uber, they are able to estimate the roughness of every road in America and precisely estimate the value people place on it, and so much more. 1/ First, what makes this all possible — acceleration data. Uber wants to know if people are suddenly braking. Their measurements along a horizontal axis incidentally allows them to measure up and down motion.