Update the writing. I found several typos without looking for them, including: "gobal" "bewteen"
Consider the following issue with your study's strategy, and solve or ameliorate them.
It seems like your main strategy is to compare "coastal cities" to "the US overall"
("Using our results we want to look at the difference in listing prices changes between the average U.S. home listing price and the coastal U.S. home listing prices to see how changes in sea levels affected the two groups.")
Your model, per your writing, seems like:
$$houseprice_{city,t} = a + bSeaLevel_t + cSeaLevelt *CoastalYN{city}$$
However, sea levels have basically been rising linearly, so this is akin to regressing
$$houseprice_{city,t} = a + bYear_t + cYeart*CoastalYN{city}$$
While the change in sea level should be worse for coastal cities than others (implying c<0), the model you've set up can show c>0 easily if other factors have overpowered the sea level rise to drive up coastal city values during your sample period.
This is why the paper I listed on the Teams idea sheet builds a property level dataset: They can design regressions that compare a house near the shore with a similar house just a short distance away except higher up. Sadly, that property level data isn't available anymore.
Getting data on sale prices at the property level might be hard, but try! If that's not available, a city has many neighborhoods at sea level and many above. The neighborhoods will be impacted differently. Can you find info on housing values at a more granular level than city? I know zipcode data is available, I found that in 15 seconds. (Neighborhood, census tract, zip code, etc)
Nice proposal.
In your revision, please
Update the writing. I found several typos without looking for them, including: "gobal" "bewteen"
Consider the following issue with your study's strategy, and solve or ameliorate them.
It seems like your main strategy is to compare "coastal cities" to "the US overall"
$$houseprice_{city,t} = a + bSeaLevel_t + cSeaLevelt *CoastalYN{city}$$
However, sea levels have basically been rising linearly, so this is akin to regressing
$$houseprice_{city,t} = a + bYear_t + cYeart*CoastalYN{city}$$
While the change in sea level should be worse for coastal cities than others (implying c<0), the model you've set up can show c>0 easily if other factors have overpowered the sea level rise to drive up coastal city values during your sample period.
This is why the paper I listed on the Teams idea sheet builds a property level dataset: They can design regressions that compare a house near the shore with a similar house just a short distance away except higher up. Sadly, that property level data isn't available anymore.
Getting data on sale prices at the property level might be hard, but try! If that's not available, a city has many neighborhoods at sea level and many above. The neighborhoods will be impacted differently. Can you find info on housing values at a more granular level than city? I know zipcode data is available, I found that in 15 seconds. (Neighborhood, census tract, zip code, etc)