Methods and data for How does Boston address its homeless problem?

Methods:

The first thing we did is use the data from the MA Emergency Assistance Shelter Report. In its Appendix A, there is a list and table of the total families. But since the report was on PDF we used an open-source tool called Tabula to get the data out from the report.

After that we could import data to the tableau and create some charts and graphs based on the family. There is a small problem when we extract these data, which is the first three towns–Acton, Amherst, Andover– their total number is on the third column instead of every town else which was on the second column. So we made a little bit of adjustment on this.

After we cleaned the data, we were finally able to import this csv file into tableau. And we got some interesting findings which we discussed for the first time. So right here when we put data into the map, we found out that a lot of the resources and facility are located on the east side of the state. Most of the families are around the greater Boston area. While the west is mostly in the Springfield region.

Another thing coming into our mind is when we read the data, we found that the money spent was much larger than we supposed. So we made a small chart about how the money changed. In this section we used the data from Commonwealth of Massachusetts Emergency Assistance Fiscal Year 2023 Second Quarterly Report and Commonwealth of Massachusetts Emergency Assistance Fiscal Year 2023 First Quarterly Report. We also made a chart of the differences between the first quarter and second quarter. But since this is not the key concept of our story, we didn’t include this into our project. There was also a problem that Tableau had to use one data at a time. The data we generated from Tabula contained three different datasets that we wanted to explore, and that makes a problem for tableau. So we have to extract the csv file into three different csv files.

Shortly after we extract this into tableau, we found another problem is that tableau has to read the date in the axis instead of column. So we went to transform the axis and column.

Since we gathered data bi-weekly, it was a great opportunity to explore how numbers changed by different cities and towns. So we went diving into the data and created these charts.

Now let’s dive deeper into the City of Boston to see how Boston addresses this problem. We also found a map of Emergency Shelter created by the city of Boston, this map includes shelters in the city of Boston. However this map does not have API or raw data provided. But they do offer to download as a jpg or PDF, so we decide to embed the photo to our story, and the photo will be hyperlinked to the website.

We also got data coming from the City of Boston Annual Homeless Census. We group them together and found an interesting thing that the total population will drop in 2021 right after the pandemic. The census is published yearly and we gathered data from 2016 to draw the following chart by datawrapper. We get the data out from the census and put it into an excel table, clean it up, and import it to datawrapper.

We made this graph to show how aggressive and effective city planning dropped the homeless population. We thought it was important to show how dramatic the drop was – something data visualization can do for stories is to make it easier for anyone to fully grasp and appreciate hard to grasp things. We also wanted to show how rapidly the population is rising from 2023 on, which is what our whole story is based on.

This is a map showing certain shelters in the Boston area that are wet (blue) and dry (red) shelters in Boston and the Greater Boston area. We found data on this from a 2017 “Directory of Resources for City of Boston” from the Volunteer Lawyers project.

We thought this would be another important part of our story, as wet and dry shelters mirror the differing and controversial methods that some shelters use. We thought it was interesting how shelters ultimately trying to do the same thing could have such a different take on how to solve the same problem – this kind of inconsistent infrastructure could be why this issue of addiction has been so hard to solve. The map graph was made with Map Customizer.

Much of the data in our story came from city reports. There is not much raw data on the homeless population, possibly because it is very hard to gather data on a group of transient people over a period of time. An invaluable resource was the Boston Public Health Division’s Unhoused and Uncounted report done in 2023. The report went deeper and had much more informative data, including but not limited to data such as income, mental health, drug addiction, and sleep deprivation of Boston’s unhoused population.

This is just some of the information we were able to gather from this report. Seeing the amount of drugs, specifically a third of the unhoused population using opiods, showed how serious the problem was and was something we wanted to point out in our story. This report, as well as data from the National Alliance to End Homelessness, made us want to talk about addiction in our story because it was clear how tied together the two issues are.

Our project also includes references to the National Alliance to End Homelessness, as well as the National Coalition for the Homeless. We also analyzed Mayor Wu’s “A Citywide Plan to Address Homelessness, Substance Use, and Mental Health”, published soon after she took office in 2021. Building off this plan, we looked at Oregon to see an example of some of the city’s methods in practice, finding some more story and data on how Portland’s homeless population has increased rapidly in recent years.

Overall, our story largely relied on

Download our data from:

Send Comment Edit Comment