Where Will We Live?

With my household's current lease expiring and one housemate leaving our household to live with his partner at the end of May, I have been worrying about where we three remaining housemates will live. This will be my seventh time searching for a new apartment to rent in the Boston area; between the three of us, we have moved within the area eleven total times in the last five years. So it's safe to say that we have been around the block, so to speak, with Bostonian apartment hunting.

It's safe to say that we have been around the block, so to speak, with Bostonian apartment hunting.

One thing I have learned from all these moves is that, unless you are moving at the same time as most of the city (Sept. 1), listings for your desired move-in date will not be posted until 1-2 months before that date. Despite looking for as much as six months in advance, my household has never landed an apartment more than few weeks before move-in.

Many realty websites have bookmarking and mapping apps, but these are limited to the listings on their specific website. Geographic Information Systems (GIS) has the capacity to organize and map the housing data while allowing for significant personalization for specific situations. To alleviate some of my housing anxiety while waiting for relevant listings to become available, I decided to use GIS to explore our housing options from a different perspective and hopefully identify affordable neighborhoods that we might have otherwise overlooked.

Ally (she/them): Val's housemate for last three years, artist, works at Trader Joe's, needs to be within a 30 minute commute (walk or transit) of her job.

Danny (they/them): Val's spouse, artist and musician, works at Artist&Craftsman, needs to be within a 45 minute commute (walk or transit) of their job.

Val (they/them): Danny's spouse, student, attends Lesley University, needs to be within a 20 minute drive of Lesley's Porter Square Campus.

Budget: We currently pay $650 a person per month for our apartment. If our incomes remain the same, we cannot afford to pay any more than $735 a person per month.

Location: Our next apartment has to be within a 30-minute commute of Trader Joe's, 45-minute commute of Artist&Craftsman, and 20-minute drive of Lesley's Porter Square Campus. Only Val/Danny have a car, so the area needs to be walkable enough that Ally is not completely reliant on Val for transportation.

Floor Plan: 2.5-3 bedrooms, 1 bathroom. While Val and Danny will share a bedroom, they both need separate work spaces because of very different noise tolerances. This could either mean two bedrooms or a bedroom and an office/dining room. Ally will have her own bedroom. Additionally, either the living room or Ally's bedroom will need to be big enough for Ally's sewing table and drafting table.

1. Identify areas in which we are most likely to be able to afford an apartment.
   1. I used  this dataset  of Median Monthly Rents by Number of Bedrooms by Census Tract for 2016-2020. I chose to use median rents because these are less likely to be skewed by outliers such as luxury apartments. The dataset was created from the American Community Survey and is the most recent rental data connected to census tracts that is available from the U.S. Census. I wanted my rent data to be organized at the census tract level in case I later decided to add crime or population data that also has a census tract field. Unfortunately, because the data was collected before the global COVID-19 Pandemic, it does not reflect changes in the rental market that occurred as a result of the pandemic. I cleaned up the CSV median rent file in Excel so that it only included the fields I cared most about: census tract (GeoID), 2 bedroom rent, and 3 bedroom rent. While I only ended up using 3 bedroom rent for my analysis, I kept 2 bedroom rent in case I wished to consider at 2 bedroom apartments (that might potentially have a suitable office/dining room) at a future time.
   2. Originally, I planned join the Median Monthly Rent CSV table to census tract shapefiles for each of the different cities within the Boston metro area that I thought were most likely to be relevant for my housing search. However, because I found that the metropolitan city boundaries did not fit neatly with the census tracts and also realized that using those boundaries did not add any information that would not already be made apparent by the reference basemap, I decided not to limit the extent of my map to only the cities I had selected at first. One of my goals for this project was to identify new potential areas, and that meant not limiting my search too early.
   3. In order to visually highlight the census tracts with median monthly 3-bedroom rents within our budget, I set manual intervals to color-code census tracts using the rent3br field by "Too Expensive" (red), "Above Current Rent" (yellow), "Below Current Rent" (green), and "Unrealistically Low" (blue-grey). I included the last category because I noticed that quite a few census tracts had median monthly rents that were unrealistically low for the Boston area (i.e. $300 for a 3-bedroom apartment) or even $0. I determined the upper bound of "Unrealistically Low" based on my experience with renting in this area, settling on $1200 as the lowest reasonable median rent. I did not want to exclude those tracts completely because it is possible that the true median rent is within our price range, but I also did not want those tracts to cloud the picture of what we definitely could afford. For ease of reading, I made the "Too Expensive" and "Unrealistically Low" tracts transparent on this storymap.
2. Identify which affordable census tracts meet our proximity requirements.
   1. The first approach I took to representing our proximity requirements was to add a MBTA train line and station shapefile to my map, since all our frequent destinations are near MBTA stations. I created a 1km distance buffer from each of the stations near our frequent destinations. This quickly showed that there was no overlap between the buffers.
   2. Then I discovered that I could run drive- and walk-time analyses on AGOL for less than 5 credits. Instead of trying to figure out appropriate distance for each buffer based on average walk speed and our desired commute times, I used drive- and walk-time analyses to create more accurate buffers around our actual destinations. I even made two levels of walk-time buffers for Danny's destination, as they are willing to walk up to 45 minutes but would prefer to walk closer to 30 minutes. I surprised by the large size of the drive-time buffer for 20 minutes to Lesley's Porter Square Campus, despite having set the traffic level to the morning rush hour. This is useful knowledge because it means that I can be very flexible with location while meeting my commute needs.
3. After completing Steps 1 and 2, I added a few more features to my map. I used GoogleMaps to find the GPS coordinates to our frequent destinations and added these as points with icons. I used the same process to add all of the places that Danny and I have lived in the Boston area to provide some familiar housing landmarks. This was valuable because several of the places we have lived are in census tracts that my map marks in red as "Too Expensive," meaning that it is possible to find apartments within our budget in those areas, just not as likely as in yellow or green areas. This makes sense because median rent is a measure of central tendency, not the lowest possible rent.

Using this map, my housemates and I have been able to pinpoint some census tracts that are the most desirable for our next apartment, so we can concentrate our energy on looking in those neighborhoods once May/June listings start to show up. We were pleased to see that there is some overlap between Danny and Ally's commute needs and that there may be 3 bedroom apartments in our budget that are comfortably within our commute needs -- it is (at least) possible!

Spring Hill, Prospect Hill, parts of Winter Hill, East Somerville, and a little bit of East Cambridge are all neighborhoods that we will prioritize in our apartment-hunting efforts.

The biggest takeaway I have from this GIS mapping experience is a little more hope for our housing future and a sense of empowerment that I have tools and skills to use spatial data to answer important everyday questions.

Final Project for Mapping Our World, Bahare Sanaie-Movahed, Lesley University, Fall 2022