Skip to main content
v2026.11,986 entries · CC-BY 4.0

Gentrification and Displacement: Definitions, Measurement and Evidence

How researchers define and measure gentrification and displacement, with a worked eviction-share example, interpretation table and APA reporting.

Written and maintained by CASRAI Editorial Board

Published

On this page

Gentrification is a neighborhood change in which higher-income, higher-educated residents move into a historically marginalized neighborhood, raising housing costs and transforming the area. Displacement is a different thing: residents being forced out, or priced out, of the place they live. The two overlap, but they are measured differently, and large-scale court-record data show that displacement through eviction is far more common in stable low-income neighborhoods than in gentrifying ones.

Definition: gentrification and displacement in one paragraph

Researchers use at least two working definitions of gentrification. The National Low Income Housing Coalition (NLIHC) reports that anti-displacement activists define it as “a profit-driven, race, and class change of a historically disinvested neighborhood”, while a narrow research definition, used by Ellen and Torrats-Espinosa, treats it as an increase of income in a neighborhood compared with the larger metro region. The University of Texas at Austin Gentrification Project describes it as higher-income and higher-educated residents moving into a historically marginalized neighborhood. Displacement is the loss of a home or a place in a neighborhood, and the same sources separate it into types.

Types of displacement

  • Direct displacement. Residents can no longer afford to remain in their homes because of rising housing costs, or are pushed out through non-renewals, evictions, eminent domain or unlivable conditions (UT Austin Gentrification Project). NLIHC describes it as residents forced to move because of rent increases or renovations.
  • Exclusionary (indirect) displacement. Low-income residents leave and other low-income households cannot move in because prices have climbed (UT Austin). NLIHC describes it as limited housing choices for low-income residents.
  • Displacement pressures. The services that low-income families rely on disappear from the neighborhood (NLIHC).
  • Cultural displacement. Local businesses and services shift toward newcomers, the character of the neighborhood changes, and longtime residents may feel dislocated even if they stay (UT Austin).

Is there a formula for gentrification or displacement?

There is no single formula. Each study builds its own classification, and the choice drives the result. A typical study first defines a unit of analysis (usually the census tract or neighborhood), then flags tracts that were low-income at a baseline year and changed on an indicator such as income relative to the metro area, then compares a displacement outcome across the flagged and unflagged groups. The outcome is the hard part, because moves are rarely recorded with a reason. The study discussed below used court-record eviction judgments against tenants as its documented measure of displacement.

For comparing two groups, the quantity used in this guide is the eviction-share ratio:

ratio = (share of all tracked evictions in a neighborhood class, %) ÷ (share of all neighborhoods that are in that class, %)

The numerator is the percentage of evictions found in that class. The denominator is the percentage of neighborhoods in that class. A ratio of 1 means the class holds evictions in proportion to its size, above 1 means over-representation, below 1 under-representation. This is a simple descriptive ratio computed from published shares, not a statistic reported by the original authors.

Worked example with real numbers: how concentrated are evictions?

Hepburn, Louis and Desmond analysed court records of 6,007,475 eviction cases from 2000 to 2016 in 72 of the 200 largest U.S. metropolitan areas. Across those cities, about 13% of neighborhoods were classified as gentrifying, and they saw nearly 12% of the tracked evictions. The 45% of neighborhoods classified as non-gentrifying low-socioeconomic-status (low-SES) saw over 60% of evictions. The data table further down this page sets these figures out.

  1. Gentrifying neighborhoods: 12 ÷ 13 = 0.92. They hold slightly fewer evictions than their share of neighborhoods.
  2. Non-gentrifying low-SES neighborhoods: 60 ÷ 45 = 1.33. They hold about a third more evictions than their share of neighborhoods.
  3. Relative concentration: 1.33 ÷ 0.92 = 1.44. Because the published shares are approximate (“nearly 12%”, “over 60%”), read this as roughly 1.4, not as a precise estimate.

The same IRP summary reports a second contrast, expressed as household risk: 1 in 16 renter households face eviction each year in majority-Black neighborhoods, against 1 in 38 in gentrifying neighborhoods. Converting: 1 ÷ 16 = 6.25% and 1 ÷ 38 = 2.63%, a ratio of 6.25 ÷ 2.63 = 2.4. In Tucson, Arizona, the median eviction rate was 5.85% in low-SES neighborhoods and 1.55% in gentrifying ones: 5.85 ÷ 1.55 = 3.77.

Interpretation: what the results do and do not show

The authors conclude that eviction acts as a stronger and more consistent driver of forced displacement in low-SES neighborhoods than in gentrifying neighborhoods, and that displacement is common while gentrification is relatively rare. The table gives reading guidance for the numbers above.

Result Reading Caution
Eviction-share ratio near or below 1 in gentrifying neighborhoods (0.92) Evictions are not concentrated there relative to their number. Says nothing about exclusionary or cultural displacement, which court records do not capture.
Ratio above 1 in low-SES non-gentrifying neighborhoods (1.33) Forced displacement through eviction is concentrated in these areas. Shares are approximate (“over 60%”), so the ratio is approximate.
Six of seven metro areas show higher median eviction rates in low-SES neighborhoods The direction is consistent across most places studied. The authors caution against broad conclusions from single-site studies, because metros differ.
1 in 16 versus 1 in 38 households per year Annual eviction risk for renters differs by neighborhood type. It describes renters only; the study does not address homeowners.

NLIHC makes a complementary point from the policy side: it says the research on gentrification and displacement is unclear, that data on displacement due to gentrification are insufficient, and that displaced low-income households most likely end up in new low-income neighborhoods.

Assumptions and common mistakes

Treating gentrification and displacement as the same variable

A neighborhood can gentrify without much measurable displacement, and displacement can occur where no gentrification is happening. The IRP authors note that displacement models developed in recent decades often center on gentrification alone and argue that they must explore a broader range of phenomena. Keep the two constructs separate in the research question, the measures and the conclusions (see operationalization in research).

Using eviction as the whole of displacement

The eviction measure covers court-documented forced moves. It misses informal moves and people who leave under rent pressure without a filing, and the study itself warns that eviction data vary by location and timeframe and that causes of eviction can be hard to determine.

Choosing the unit of analysis carelessly

Neighborhood-level patterns do not describe individual households. Inferring individual experience from area-level shares is the ecological fallacy, covered in ecological study design and the ecological fallacy, and the choice of areal unit is discussed in unit of analysis. Methods for neighboring-area dependence are in spatial data analysis.

Generalising from one place or period

The study covers 72 U.S. cities for 2000 to 2016 and relationships may differ at other times and in other locations. A purposive sample is not a random sample; see sampling bias.

Ignoring who is vulnerable

The UT Austin project names renters, low-income households, persons of color, households headed by someone without a college degree and families with children in poverty as more vulnerable to displacement. Studies that look only at neighborhood averages can hide who bears the risk.

How to report gentrification and displacement results (APA example)

State the classification rule, the baseline and end years, the unit of analysis, the displacement measure and its data source, then give the descriptive shares. An APA 7th edition style example for the figures above:

In 72 large U.S. metropolitan areas from 2000 to 2016, about 13% of neighborhoods were classified as gentrifying and accounted for nearly 12% of tracked eviction cases, whereas the 45% classified as non-gentrifying low-SES accounted for over 60% (Hepburn et al., 2024).

Reference entry:

Hepburn, P., Louis, R., & Desmond, M. (2024). Beyond gentrification: Housing loss, poverty, and the geography of displacement. Social Forces, 102(3), 880–901. https://doi.org/10.1093/sf/soad123

Name the definition you adopt (for example the narrow income-relative-to-metro definition) and say that other definitions exist, since the label determines which tracts are counted. For Census-sourced covariates see how to cite Census data.

How to calculate the eviction-share ratio in Excel, R or Python

The logic is the same in each tool: divide the eviction share by the neighborhood share. With the shares above, each returns 0.92 for gentrifying and 1.33 for low-SES neighborhoods.

Excel

Put the neighborhood share in B2 and the eviction share in C2, then in D2 enter =C2/B2. With B2 = 13 and C2 = 12 the result is 0.923; with B3 = 45 and C3 = 60, =C3/B3 returns 1.333.

R

nb  <- c(gentrifying = 13, low_ses = 45)   # % of neighborhoodsev  <- c(gentrifying = 12, low_ses = 60)   # % of tracked evictionsround(ev / nb, 2)                          # 0.92 1.33round((1/16) / (1/38), 1)                  # 2.4

Python

nb = {"gentrifying": 13, "low_ses": 45}ev = {"gentrifying": 12, "low_ses": 60}print({k: round(ev[k] / nb[k], 2) for k in nb})   # {'gentrifying': 0.92, 'low_ses': 1.33}print(round((1 / 16) / (1 / 38), 1))             # 2.4

With real tract data, compute the shares from counts (evictions in class ÷ all evictions; tracts in class ÷ all tracts) rather than typing percentages, and report the denominators.

Where this fits in research methods

Studies of gentrification and displacement combine a construct that must be operationalized, an areal unit, a documented outcome and a comparison group. Neighborhood housing studies often draw on household surveys such as the American Housing Survey, and the topic sits within the wider field of geography and urban research.

Sources

Data: Eviction concentration by neighborhood class in 72 large U.S. metros, 2000-2016

Shares of neighborhoods and of tracked eviction cases by neighborhood class, from the Hepburn, Louis and Desmond (2024) study summarised by the Institute for Research on Poverty, with the eviction-share ratio computed here as eviction share divided by neighborhood share, and the reported annual renter eviction risk as 1 in N households. Shares are the published approximations (nearly 12 percent, over 60 percent).

Period covered: 2000-2016. Coverage: 72 of the 200 largest U.S. metropolitan areas.

Eviction concentration by neighborhood class in 72 large U.S. metros, 2000-2016
Neighborhood classShare of neighborhoods (%)Share of tracked evictions (%)Eviction-share ratioAnnual renter eviction risk (1 in N households)
Gentrifying13120.9238
Non-gentrifying low-SES45601.33not reported
Majority-Blacknot reportednot reportednot computable16

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Gentrification and Displacement: Definitions, Measurement and Evidence

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →