Introduction

This project is motivated by the question of whether certain boroughs in NYC are more prone to homelessness (measured by encampments) and drug usage than others, and whether this has changed over time. This question is important to policy makers and community advocates who may want to demonstrate to the world that there is economic segregation within NYC’s boroughs.

To answer this question, I leverage 311 data from the City of New York, between 2016 and 2025. This dataset consists of complaints that New Yorkers called in or posted online to the City, and can range from noise complaints to illegal parking and apartment maintenance issues. 

From this dataset, my variables of interest include years; the number of encampments; the number of reports of drug activity; and the boroughs where each complaint occurred. 

I hypothesize that the Bronx, Brooklyn, and Queens will have the highest levels of encampments and drug usage because these are boroughs with high concentrations of poverty. I believe, by contrast, that Manhattan and Staten Island will have lower numbers of encampments and drug usage because they have lower poverty. Finally, I anticipate that encampments and drug usage will be highest between 2020 and 2021 because that was the peak of the COVID-19 pandemic and people were struggling financially and emotionally.

Notably, this dataset relies on complaints, and is therefore not a perfect proxy for actual drug activity or homelessness. People may perceive drug activity or homeless encampments where there are none. Furthermore, the drug activity or encampment had to bother someone enough to complain–suggesting that this dataset may underreport actual levels of drug activity and homelessness in NYC’s five boroughs.

Figure 1. Pie Chart

I first sought to understand the overall landscape of drug usage and encampment complaints across New York City. This pie chart shows that, between 2016 to 2025, there were 3 drug complaints for every encampment complaint in this period. In terms of sample size, there were 26,943 complaints regarding encampments and 69,983 complaints regarding perceived drug usage. While seemingly large, these numbers strike me as small given that these complaints are supposed to span a decade, and New York City is a very large place.

Figure 2. Bar Graph (Small Multiples)

Having understood more about the overall number of complaints, I investigated the distribution of complaints by borough over this 10-year period. The small multiples bar graph allows the depiction of each borough’s drug complaints alongside encampment complaints.

The figure shows that most drug activity is reported in Queens, whereas most encampments are reported in Manhattan. We cannot infer whether there were more encampments and drug usage actually occurring in these boroughs given that we are relying on complaints data. However, the data creates a hypothesis that can be tested by further research using a more reliable dataset.

Interestingly, this graph also indicates that drug usage and encampments don’t necessarily covary with one another. That is, the borough with the most encampment complaints (Manhattan) doesn’t also have the highest reported drug activity in NYC, and vice versa.

Figure 3. Pie Chart (Small Multiples)

This small multiples chart depicts the same information as Figure 2, except pie charts are used instead of bars to give the reader a sense of the relative proportions of complaints in each borough. Looking at the data in this format, I realized that Staten Island and the Bronx actually had remarkably low numbers of drug and encampment complaints relative to the other boroughs. This counters my hypothesis that the Bronx would have high incidences of both complaints because of high poverty levels.

Figures 4 and 5. Time series (Line graphs)

Figures 4 and 5 are meant to be read together because they both encode drug usage complaints between 2016 and 2025. Notably, Figure 4 shows that four of the five boroughs in NYC tend to have constant levels of drug activity over time. Although some lines fluctuate by a few thousand (e.g., in Manhattan, Brooklyn and the Bronx), the overall trend is pretty stagnant. 

Figure 5, however, depicts Queens all on its own to demonstrate that there were greater fluctuations in drug usage in this borough than the other boroughs. Between 2021-2024, alone, drug usage reports spiked by 17,000. It then dips in 2025 (likely because this dataset was collected in the first quarter of 2025, so it does not reflect a full year of potential drug usage complaints). 

I am surprised that drug activity spiked so much in Queens relative to the other boroughs. It could be due to increasing poverty, the rising cost of living, and general social despair since the pandemic. However, it is a mystery that drug activity didn’t rise for these same reasons in the other boroughs. This may be an artifact of how our data was collected–i.e., people in Queens may be more likely to complain about perceived drug usage than people in other boroughs. We can’t know for sure.

Figure 6. Time series (Line graph)

Discussion and Conclusion

I hypothesized that homelessness (measured by encampments) and drug usage complaints would be highest in what I considered to be the poorest boroughs of the city–Queens, the Bronx and Brooklyn. By contrast, I figured that Manhattan and Staten Island would have less complaints because of what I perceived to be better social conditions.

While the data indicate that Queens has the most drug usage complaints, the most homelessness is actually reported in Manhattan. Meanwhile, Staten Island, the Bronx and Brooklyn tend to report lower levels of each complaint. Furthermore, we see that drug usage stays pretty stable over time in all boroughs except Queens, and encampments are only reported as a problem at the peak of the pandemic. 

Ultimately, I believe the way this data was collected limits its ability to answer the questions I set out to answer in this project. People had to be compelled to report a problem, which means that likely not every instance of drug usage or encampments was reported. Future research should find a more reliable data set to operationalize homelessness and drug usage in New York City. This will provide a clearer picture of quality of life in the five boroughs.