New blog
I’m continuing my blog over at Substack: https://blog.michaelwiebe.com/ . Check it out!
The blog of Michael Wiebe
Latest posts
Condonomics Debunked: A review of Patrick Condon's 'Broken City'
In Broken City , landscape architect Patrick Condon presents a diagnosis of the housing crisis. Condon claims to be a pragmatic empiricist, but he actually relies on extreme theoretical scenarios. He thinks in memes, not models. The result is a flawed analysis of housing policy. Condon misunderstands the relationship between land and housing prices, and incorrectly identifies upzoning as the main…
A review of Murray's 'Great Housing Hijack'
In The Great Housing Hijack , Cameron Murray sets up a framework with five equilibria (asset price, rental, spatial, density, and absorption rate), which he applies to various policy issues. I found a couple of insightful points. For example, the price-to-income ratio is a flawed measure of housing (ownership) affordability, because we really care about the cost of housing services (i.e., mortgage…
A two-sector model of land and upzoning
When we upzone land from single-family zoning to apartment zoning, we change the allocation of a city’s fixed stock of land. Upzoning makes apartment-zoned land more abundant and hence cheaper, while single-family-zoned land becomes scarcer and more expensive. Because land is an input into the production of housing, upzoning reduces the cost of building apartments, and hence is a key policy option…
A YIMBY FAQ
What caused the housing crisis? Skyrocketing housing costs in big cities are caused by demand for housing increasing faster than supply. The key mechanism underlying the housing crisis is the demand cascade : richer people move in and bid up the price of new homes, which pushes less-rich locals to compete for old homes, which forces the poor to take on roommates, move away, or become homeless.…
A supply and demand model of housing, part 2: unit demand
In the last post I used continuous quantities of housing: people could buy, say, 1.26 homes. Here I show how a supply and demand model works with discrete quantities, and with a unit demand constraint, where people buy at most one unit of housing. This involves a different approach than the perfect substitutes case, but we arrive at the same conclusion regarding vacancy chains, demand cascades,…
A supply and demand model of housing, part 1: continuous quantities
When we build new apartments, the people moving in vacate their old apartments; this reduces competition for old housing, making it more affordable. This is the vacancy chain effect: building expensive new housing helps improve affordability of cheaper old housing. When rich people move into a city, they outcompete locals for new homes; those locals in turn compete for the stock of old homes,…
Notes on the new measles literature
There are a handful of papers applying the ‘disease burden’ method to study the long-term effects of the measles vaccine. This method uses cross-sectional pre-treatment disease incidence interacted with a time series variable for vaccine access. This works for diseases like hookworm and malaria, which have geographic variation in climatic suitability for the parasites that cause disease. But, as I…
Summary: replication of Moretti (2021)
Moretti (2021) is about agglomeration effects in innovation: do inventors patent more when they’re around other inventors? In other words, does the size of tech clusters cause patenting? This question is relevant for housing policy, because high housing costs prevent inventors from congregating in tech clusters. So if agglomeration effects are large, then constraints on housing supply are…
Can we detect the effects of racial violence on patenting? Replicating Cook (2014)
A year ago, I wrote a short post looking at the data in Cook (2014) ( sci-hub ) ( replication files ) on the effect of racial violence on African American patents over 1870-1940. I discovered that the state-level panel data was strikingly imbalanced. With Lisa Cook in the news for being nominated to the Federal Reserve Board of Governors, I decided to revisit the paper more thoroughly. I find that…
Did medical marijuana legalization reduce crime? A replication exercise
Summary In this post I replicate the paper “Is Legal Pot Crippling Mexican Drug Trafficking Organisations? The Effect of Medical Marijuana Laws on US Crime” by Gavrilova, Kamada, and Zoutman (Economic Journal, 2019; replication files ). I find three main problems in the paper: it uses weighting when its own justification doesn’t apply it uses a level dependent variable, and isn’t robust to…
How I use regression weights to replicate research
One of the main tools I use for replication is regression weights . These show the weight that each observation contributes to a regression coefficient. Suppose we’re regressing \(y\) on \(X_{1}\) and \(X_{2}\), with corresponding coefficients \(\beta_{1}\) and \(\beta_{2}\). Then, the regression weights for \(\beta_{1}\) are the residuals from regressing \(X_{1}\) on \(X_{2}\), which represent…
Does meritocratic promotion explain China's growth?
One explanation for China’s rapid economic growth is meritocratic promotion, where politicians with higher GDP growth are rewarded with promotion. In this system, politicians compete against each other in ‘promotion tournaments’ where the highest growth rate wins. This competition incentivizes politicians to grow the economy, and hence helps explain the stunning economic rise of China. The…
Replicating the literature on meritocratic promotion in China
China has had double-digit economic growth for nearly three decades. How can we explain this? In my dissertation, I studied one explanation that is backed up by a large literature: meritocratic promotion. The idea is that politicians compete in promotion tournaments, where the politician with the highest GDP growth rate in their jurisdiction is rewarded by being promoted. By tying promotion to…
Dropping 1% of the data kills false positives
How robust are false positives to dropping 1% of your sample? Turns out, not at all. Rachael Meager and co-authors have a paper with a new robustness metric based on dropping a small fraction of the sample. It’s called the Approximate Maximum Influence Perturbation (AMIP). Basically, their algorithm finds the observations that, when dropped, have the biggest influence on an estimate. It calculates…
Is randomization inference a robustness check? For what?
I’ve seen a few papers that use randomization inference as a robustness check. They permute their treatment variable many times, and estimate their model for each permutation, producing a null distribution of estimates. From this null distribution we can calculate a randomization inference (RI) p-value as the fraction of estimates that are more extreme than the original estimate. (This works…
How to p-hack a robust result
Economists want to show that our results are robust, like in Table 1 below: Column 1 contains the baseline model, with no covariates, and Column 2 controls for \(z\). Because the coefficient on \(X\) is stable and significant across columns, we say that our result is robust. The twist: I p-hacked this result, using data where the true effect of \(X\) is zero. In this post, I show that it can be…
Why we shouldn't take p-values literally
PSA: if you read a paper claiming p<0.01, you shouldn’t automatically take that p-value literally. By definition, a p-value is the probability of getting a result at least as extreme as the one in your sample, assuming the null hypothesis H0 is true. In other words, assuming H0 is true, if you collected many more samples, and ran the same testing procedure, you’d expect to find that p% of the…