Some good coverage of my working paper on housing filtering in The Atlantic today.
The paper itself, The Downmarket Impact of New Multifamily Housing: Evidence from a Honolulu Condo Tower, is available here: papers.ssrn.com/sol3/papers....
Using detailed address-level microdata, we trace movement chains originating with the initial residents of The Central and document three main findings. First, we show the building generated a substantial number of local vacancies. We identify 180 specific addresses that became vacant because of moves into The Central. Scaling to account for data coverage suggests the new tower induced more than 500 local vacancies in the three years after construction, by setting off chains of moves. Second, while The Central units were expensive on a per-square-foot basis, the homes vacated by movers were significantly cheaper. Homes left behind by those moving into The Central were about 40% less expensive. Unlike much of the prior literature, which track the changing neighborhood characteristics of movers
Even more evidence that building new housing decreases rents: Researchers tracked the residents of a newly built luxury condo building and found that they freed up less expensive apartments nearby.
papers.ssrn.com/sol3/papers....
Want to live and work in this beautiful mountain valley in HawaiΚ»i? #econsky
The University of Hawaii Economic Research Organization (UHERO) is hiring for 2 TT positions:
-> Public Finance/Tax: aeaweb.org/joe/listing....
-> Macro/tourism/forecasting: aeaweb.org/joe/listing....
More below
1/5
New working paper posted on recent US experiences with fare-free transit service:
papers.ssrn.com/sol3/papers....
π¨ We're hiring for 2 tenure track Economics positions at the University of Hawai'iπ¨
βοΈ ποΈ π΄ π€
#econsky
1) www.aeaweb.org/joe/listing....
2) www.aeaweb.org/joe/listing....
With nighttime pedestrian deaths on the rise, new research on the link between moonlight and pedestrian fatalities suggests that even minimal lighting interventions to dangerous roads can save lives.
nextcity.org/urbanist-new...
Non-paywalled version of the article can be found here: www.justintyndall.com/tyndall_moon... π
Moonlight and artificial light might offer different safety benefits. Streetlights provide ~30Γ more illumination than the brightest moon. My estimates probably understate the impact of full street lighting. Even very small increases in road illumination significantly improve pedestrian safety.
If roads were always lit to at least the level of the brightest moon, 9% of all pedestrian deaths could be prevented. Investing in lighting along dangerous roads has huge returnsβone Florida case shows an ROI of 4,000%.
I find that pedestrian deaths fall 5% nationally when the moon is at its brightest. In areas without cloud cover, the effect is 17%. In rural areas with low artificial lighting, the effect is nearly 40%!
I analyze 48 years of US crash data (287,023 pedestrian deaths). For every night, I calculate the number of pedestrians who died and the moonβs brightness. I compile satellite data on cloud conditions and artificial lighting to account for their influence.
The moonβs brightness changes nightly due to its phase, distance, and other factors. Our calendar is not set by the moonβs phase, so across years the moonβs brightness is changing randomly with respect to the day of the year or week. I use this as a natural experiment.
To test lightingβs impact, weβd need a randomized experiment. But turning streetlights on/off for research would be unethical. Howeverβ¦ nature changes road illumination every night via THE MOON.
The night is more dangerous for pedestrians because visibility is low, but also due to driver fatigue, worse weather conditions, more drunk driving, etc. So how much does lighting matter?
Nighttime pedestrian deaths have DOUBLED since 2009. Understanding the causes of nighttime pedestrian deaths has become increasingly important.
My paper, Road Illumination and Nighttime Pedestrian Deaths: Evidence from Moonlight, is now published at Economics of Transportation. authors.elsevier.com/a/1kptw_oIvi...
I analyze 48 years of US crash data (287,023 pedestrian deaths). For every night, I calculate the number of pedestrians who died and the moonβs brightness. I compile satellite data on cloud conditions and artificial lighting to account for their influence.
The moonβs brightness changes nightly due to its phase, distance, and other factors. Our calendar is not set by the moonβs phase, so across years the moonβs brightness is changing randomly with respect to the day of the year or week. I use this as a natural experiment.
To test lightingβs impact, weβd need a randomized experiment. But turning streetlights on/off for research would be unethical. Howeverβ¦ nature changes road illumination every night via THE MOON.
The night is more dangerous for pedestrians because visibility is low, but also due to driver fatigue, worse weather conditions, more drunk driving, etc. So how much does lighting matter?
Nighttime pedestrian deaths have DOUBLED since 2009. Understanding the causes of nighttime pedestrian deaths has become increasingly important.
Our latest investigation with @propublica.org found that FEMA sidestepped its own policies when setting up a housing program for Maui wildfire survivors. As a result, some people were pushed out of their homes by landlords seeking higher rents. www.civilbeat.org/2024/11/maui...
Honolulu is considering an Empty Homes Tax. ποΈ
I crunched some numbers on the possible impacts on available local housing and county revenue.
There's significant uncertainty, but the potential benefits are large.
uhero.hawaii.edu/a-honolulu-e...
π«‘
www.whitehouse.gov/wp-content/u...
Excellent summary of my paper by a Canadian motorsports company β¬οΈ
www.youtube.com/watch?v=YpuX...
A non-paywalled version is available here: www.justintyndall.com/uploads/2/8/...
I estimate how many pedestrians would be saved if front-end vehicle heights were capped at 125 cm (about the height of a new Honda CRV). I estimate 500 fewer pedestrians would die each year if taller front-ends could be lowered to 125 cm.
I show the effect of tall vehicles is more pronounced among women, children, and seniors. The front-end effect on death probability is 4x as strong for crashes involving children.
I find a vehicleβs front-end height is more predictive of pedestrian death compared to other measures of vehicle size. A 10 cm increase in a vehicleβs front-end relates to a 22% increase in the probability a pedestrian dies, controlling for crash characteristics.
