Visualization: Gentrification, Racism, White Flight and Policing

Gentrification is an issue that involves many different factors. When looking into this, housing costs, eviction rates, development, employment opportunities, race, gender and even access to parks affected the rate of gentrification. In this project, I’m defining gentrification as the displacement of marginalized people. So far, I’m going to be using population size and house ownership to assess white flight. White flight is kind of the first step to an area being gentrified.

I’m having trouble knowing how much data I want to include and what would create a realistic and all-encompassing view of gentrification in Pittsburgh. Overall, Pittsburgh has low displacement rates when compared to the country, but that doesn’t change the fact of how harmful gentrification can be on communities, real people, and culture.

As companies like Uber and Google move in, and as they take advantage of cheap land, I wonder what the future holds for communities already struggling with making mortgage payments.

I haven’t been able to find this data but I need to find numbers on housing instability.

Here are some resources I’ve been scanning to find more relevant information. Based on what I’ve found, gentrification is something we should be aware of always, especially in the context of Pittsburgh.


Q: What happens after displacement?

A lot of what I’ve been reading centers around the predictions of gentrification or places in the middle of being gentrified. Through my investigations, I think I’m more interested in investigating what happens to an area after gentrification has gotten a hold of it. I want to focus on the policing of people of color and what happens to the facilities once they are ditched by white flight. Right now my data is in a LOT of places. I have way too many points I’m looking at so I really need to pare down. Another topic or angle I could take is environmental racism.

11/12 : Class

Reviewing Yau

Scales :

  • logrithmic ( 0, 10, 100)
  • linear (0, 1 , 2, 3) (alphabetical) (time)
  • categorical (red, blue, green)
  • time ( day/month/year, hour, linear/cyclical, seasonal) (time)
  • percentages (parts of a whole)
  • Ordinal ( spectrum; bad, neutral, good) (Hierarchy)
  • Location

Breaking things down in class
1) Write Categories
2) Then write scales onto them
3) Then put things into buckets/groups
4) If there are only 2 sides, think how you can break things down into further granularity
5) Then separate them into coordinate systems (cartesian, polar, geographic)
6) Do we start to see a story building? Can we start to organize the post-its
7) What's the starting point? Figure that out, then dive into it
8) Look at narrative and indexical structures
9) What coordinate system might be an anchor for this?

!Think about layering story! How will that pan out interactively?

After class….

I was still having trouble understanding the story I wanted to tell. I had the information I needed, I just wasn’t seeing the story yet. So after class, I took to the whiteboard to lay out the parts and pieces to try and see the relationships between them.

What did I nail down? Well, Gentrification is very intersectional. The fact leading up to, during and after involves many moving pieces.

I went home, took a break, and looked at how people were defining gentrificaiton. I found this tool that shows the 3 major methods of how people define gentrification:

There are 3 main types. Freeman, Ellen & O’Regan, McKinnish et al. They all define gentrification based on parameters they set. Personally, the way I had understood gentrification as was how Freeman & McKinnish were describing it. So I think i’ll use their parameters to define it.

I’m also personally interested in the policing of historically black neighborhoods as they become gentrified. So I did some research.

I’m not sure how Pittsburgh handles the policing of the black community but I know tensions exist between them. Rest in Power Antwon Rose✊🏽

Through this research, I want to investigate the potential relationship between policing in Pittsburgh over time and the areas that are identified as gentrified. I would like to layer on other signifiers of gentrification such as education levels, income, and average house costs.

It’s tough finding granular information for some of the data I need on each neighborhood. I also do wish it was 2021 so we had access to the 2020 census!

11/13 + 14: Data Types

Below is a breakdown of the data and data types I have aggregated thus far

I’m not sure if I want to focus on this ‘over time’ or just focus on one year. Either way. I have data on Population by Race in Neighborhoods, Median Income by Neighborhood, Police reports by Neighborhood, and Median Housing Costs by Neighborhood. I chose these because it’s a mix of data that are good indicators of a place that’s becoming gentrified. I also want to assess the relationship between gentrification and the policing of neighborhoods.

These continue down to include all neighborhoods!

For the sake of time, I think I will focus on a specific chunk of neighborhoods. I chose the following based on articles indicating which neighborhoods were socio-economically disadvantaged and which neighborhoods were being gentrified.

Neighborhood Breakdown
Gentrified + “Ones to Watch”

  1. Lawrenceville
  2. Bloomfield
  3. Garfield
  4. Polish Hill
  5. Downtown
  6. North Side
  7. Mount Washington
  8. Downtown
  9. St. Clair
  10. Mount Oliver
  11. East Liberty
  12. South Side
  13. Homewood
  14. Hill District
  15. Strip District

Organizing the Data


In class I finally pinned down my data points and made more sense out of how these point were going to work together. Yay!

Outside of class I started to sketch out visual cues for what I was doing. I first started with a word map to get me thinking of visual cues that come to me when thinking about gentrification. Lots of circles (bullet holes) and lines (uprooting, roots). Tonight I’ll try thinking about layering things but I still need to do more individual explorations. I also think I need to organize my thoughts on visual cues a little more clearly :-) but for now i’m just exploring!


  1. What question or questions are you exploring? Is there a relationship between policing and a gentrified neighborhood? Does policing go up if the neighborhood was rapidly gentrified or slowly.
  2. What types of data are you using? The following are all for each neighborhood. Median Home Value (over time), Median Household Income (over time), Neighborhood population (by race and time), # of crime reports (by year), # of 311 reports (by year), non-arrests (by year).
  3. What coordinate system are you using? I am using cartesian. I am using cartesian because I want to visualize the relationship across each neighborhood. So that relationships can be easily distinguished.
  4. What scales do you plan to use for each type of data? Median Home Value (Location), Median Household Income (Ordinal), Median Home Value (Ordinal), Neighborhood Population (Logarithmic), # of Crime Reports (Category)
  5. What ranges are you using for each type of data? Median Home Value (0–10k,10.1k-30k,30.1k-60k,60k-90k,90–140k,140+) Median Income (High, Above Average, Average, Low, Poverty) Neighborhood Population (%White, %Black), #of Crime Reports (High Medium Low)
  6. Do you propose using a narrative or indexical structure for your visualization? I think I’m going to be using an indexical structure? I little bit of a narrative at the beginning and then diving into indexical? What is the projective data, what will people see or uncover at each step?
  7. What visual, aural, or temporal cues do you plan on using? Where are you going to apply it and why? I’m definitely going to be using color to set the tone over time. I think using scale would be important to use to convey change over time. Maybe utilizing loud industrial sounds to visualize the amount of policing in a neighborhood over time? Motion will be helpful to see the change between each year, you can visually see things grow over time. I want to utilize depth to show the loss of communities but not sure how to. Density and distance for population and home value.
  8. The answers to these questions ^ are what I need to present on Tuesday.
  9. Next step is to transition to form.
  10. Why are we doing this project? Changing data into information. Information can be understood. Adding meaning. Storytelling! Helping other people see what we’re seeing. Patterns. Taking disparate parts, data from different places and helping people see what those relationships are. Helping other people see patterns that aren't easily visible. Bringing sense to dense information, how do you share that with someone else? Then also adding interaction in this project.

Notes on Class

Pattern & Detection
* Number & Detection
(people can usually perceive 7 different things aka 7 different colors, 7 different buckets)
* Temporal Building = Building something over time. What you’re making here isn’t a static piece, you have time on your side. What are you introducing first, second, third? How to bring someone into the story. This will have entry points. We’re working towards the deeper engagement. How do I leverage temporal building? How do I present something over time? What’s my starting point? (Show all neighborhoods) I think I’ll be showing socio-economic indicators first and then policing.

* Categorization
Similarities in two directions
* Pacing & Simultaneity Be mindful of where it is important to see information at the same time rather than one after the other. Where in the project do you want people to see at the same time? What do I want to show simultaneously? What do you have to see all the time?

You can show everything one a time or at the same time. Consider what’s important for the viewer to see.

Narrative & Indexical Structures

Don Norma — Things that make us smart :
Appropriateness Principle: Providing the right amount of information??
Can people grasp things quickly without having to explain it to them? You probably will bring a key (that’s ok!) in but how far can we go without using one?

Semantic Differentials
Used in design when figuring out form when there isn’t one.
Mechanical -> Free Form (also includes Organic) Loud -> Quiet
Take adjectives that are opposites and see where your project lies.

Expectations + Perception
Consider audiences' schemas or perceptions of things before they see your piece. What is your audience likely to come to this with this? Present information to make people think critically about it. Using data to understand what questions to ask.

you really have to consider the role of your audience. how much customization can the audience actually do? What options will you actually provide to people? How many options are you going to give them? What is the customization? How much choice and freedom are you going to give people?
Mimicking Known Behaviors don't base information on limited information (don't use stereotypes) There is nothing innovative about this! We need to push boundaries, the challenge is finding a sweet spot. How do you build on known behaviors but move to a different place? How do you build on what people know but push it? How do you build their curiosity and interest in this?

Recall + Engagement
How do you leverage those past experiences. How might you leverage those to encourage engagement? What is something that you want to trigger with them? Visuals should help set context.
Discovery + Critical Thinking Think about the rewarding experiences that you’ve had. Chances are they didn’t provide all of the answers for you. How much can I put in there to help people figure it out? It can come across as pandering if you don’t allow for discoverability. How do you frame your question so that people WANT to think critically about it? Instead of pointing out the obvious, just lay out the data and then people find the answers. You don’t want to tell them all the answers.

For Today : Focus on temporal building AND the representation issues AND simultaneity (layering)

What is just *one* pathway through the data?

Presentation Day

Holiday Break Update

After our presentation, I felt a lot better with the direction I was going in. But it was clear I needed to start doing more visual explorations. Below are a few examples of where my head is at

1: Thinking about using motion to show “strength” of air by using a ticker style so the neighborhoods name would move quickly through the row if the air quality is good (NO2). Buti t would move slowly and lackluster-ish if the NO2 levels were high. I gravitated towards a more swiss style.

2: Polarrrrr. I’m not sure that polar is actually right but I took a stab at it to see if it would be right. I could really visualize it in my head until I put it down on paper. I’m still trying to work through and understand the different cues I’d want to use and understand the different ‘toggles’ or interactivity I’d want to use.


AHHHHH im stuck

MDes @ CMU • They/Them • The scum between your toes