Paper: Lester, R. (2026). Rational Foreclosure: Hope as an Economic Variable. Working paper, University of Houston. Read the current version (PDF) · SSRN 6614858.
Try it: The Hope Gap shows, for any U.S. state, county or large city, how far people are from giving up on a home or a decent standard of living, and what it would take to bring them back.
Planning tool: The Hope Gap, for federal, state, county and city officials. Data refresh twice a week.
Replication package: Harvard Dataverse, doi:10.7910/DVN/V5NSZX (version 2 is the October 2026 revision).
A generation of young workers stands accused of quiet quitting: showing up, doing the job, and no longer chasing the long hours their parents treated as the price of getting ahead. The usual diagnosis is a lack of grit, and the usual cure is a push to keep going. The paper shows that the diagnosis can be backwards.
Picture someone saving for a home whose price keeps moving. She measures her progress against that price, and falling behind hurts more than getting ahead pleases. The paper proves that she keeps chasing while she is close, sacrifices less as the price pulls away, and stops keeping score at a precise point below it. That point can be calculated, and three things move it:
- Swings keep people in. If the price swings up and down, it may swing back toward her, so she rationally holds on longer. The value of that chance is hope, and the paper gives it an exact value.
- A runaway target pushes people out. If the price climbs steadily away from her pay, she quits sooner.
- Time left matters. Within a working life, everyone has a quitting point, and of two people equally far behind, the one with fewer years left stops first.
Pushing her to keep going past her quitting point makes her worse off, and the paper says by exactly how much.
Because everyone in a market faces the same prices, one shock moves a whole generation toward its quitting point at once. A runaway target becomes a disengaged workforce, and the costs follow: lost effort, homes and families put off, and the strain on health that the evidence links to staying in a losing race. The remedy lies in the target, not the worker: slow the prices that have run away, or raise the pay that chases them.
Pick a place, a goal (owning a home, or keeping up with the cost of living) and a group. The tool reports the Hope Gap for that place (the tipping point minus the gap between the goal's growth and pay growth, in points a year; above zero means hope holds), the model's quitting point, how many people are past it, what happens under several scenarios, and what it would take in homes built or pay growth to close the gap, with a dated plan for the chosen office. By default the race ends at 67, so each person's quitting point depends on the working years they have left. Each group gets its own pay trend (Census Quarterly Workforce Indicators, by age, sex, education, and race and ethnicity), the lowest income group is measured against starter homes, and a comparison table shows every group side by side. The model is solved in the browser; nothing the user enters leaves the page.
.
├── paper/ # the paper (PDF) and its LaTeX source
├── tool/ # The Hope Gap: page template, build script, and an offline copy with data built in
├── pipeline/build_data.py # builds data/core.json, counties.json, places.json from FRED, FHFA, BLS QCEW, Zillow and the Census Bureau
├── pipeline/build_groups.py # builds data/groups.json, group pay trends, from the Census Quarterly Workforce Indicators
├── .github/workflows/data.yml # runs the pipeline twice a week (group data about monthly) and publishes to the gh-pages branch under /data
├── code/ # replication code for the paper (mirror of the Dataverse package, version 2)
├── data/README.md # where the tool's data live and how they are built
└── README.md
The gh-pages branch holds the live page (index.html) and the published data (data/). An offline copy of the page, with a data snapshot built in, is in the Dataverse package (tool/the_hope_gap.html).
pip install numpy scipy matplotlib
cd code
python3 sweeps.py # Tables 2 and 3 and the table in Result 1
python3 deadline.py # Proposition 4: quitting point by years left (Table 4)
python3 welfare.py # Proposition 5 magnitudes, Result 3, Figure 2
python3 cohort.py # Table 5
python3 figs.py # Figures 1 and 2appendixA_numerical_checks.py runs the 243-point scan behind Appendix C (10 to 20 minutes). shoot.py is the independent shooting solver used by the tool; it matches the collocation solver to four digits. See code/README.md for the full table of scripts and the results each reproduces.
FRED (BLS and BEA national series), FHFA House Price Index (state and county), BLS Quarterly Census of Employment and Wages, Zillow Home Value Index, Census American Community Survey. Sources, dates and fallbacks are listed on the page under Data and methods.
@unpublished{lester2026foreclosure,
author = {Lester, Ryan},
title = {Rational Foreclosure: Hope as an Economic Variable},
year = {2026},
note = {Working paper, University of Houston. SSRN 6614858. Replication package: doi:10.7910/DVN/V5NSZX},
url = {https://github.com/Lesterhau/rational-foreclosure}
}CC BY 4.0 (code, text and data). Data published under data/ on the gh-pages branch are derived from the public sources listed above and remain subject to their terms.
Ryan Lester, rslester@cougarnet.uh.edu, ORCID 0009-0002-7840-5676.