The article reports that county-wide SNAP participation rose from 4.7% in 2019 to 6.7% in 2023.[1] A two-point increase over four years. These rates are not themselves a direct measure of food insecurity.
But what happens when we look at neighborhoods instead of the whole county?
The article reports a more complex picture: of a stated 334 census tracts with complete data, 160 are getting worse, 67 are getting better, and the rest hold steady. The denominator is under reconciliation.
It also reports a vulnerability gradient: 76.7% of the highest-vulnerability neighborhoods are deteriorating, compared with 15.9% of the lowest-vulnerability neighborhoods. That comparison has not yet been reproduced from a matched output.
Let's explore what's happening.
The article uses diverging trajectories to describe a reported split between 160 census tracts where SNAP participation increased by more than 1 percentage point from 2019 to 2023 and 67 tracts where it decreased. The reported percentages, resident count, and 2.4-to-1 ratio inherit an unresolved tract denominator and should not yet be treated as reproduced findings.
The Background
Research shows that food insecurity tends to concentrate in certain places and persist over time.[2][3] Poor neighborhoods tend to stay poor. Communities that struggled before economic shocks tend to struggle more after them.
What's less clear is whether neighborhoods are moving toward the county average or pulling apart. Are struggling areas catching up? Falling behind? Holding steady?
To find out, we classified each tract's four-year path:
- Deteriorating: SNAP participation rose more than 1 percentage point
- Stable: Changed less than 1 percentage point
- Improving: Fell more than 1 percentage point
This simple classification reveals patterns hidden in the county average.
How We Did This
Where: Santa Clara County, California (408 census tracts, 1.9 million residents)
Data: Census ACS 5-Year Estimates (2019-2023) for SNAP rates by tract, plus demographics like race, education, poverty, and vehicle access.
Approach: The article describes tracking tracts over five years, classifying their trajectories, and comparing those trajectories with vulnerability and demographic measures.
Reconciliation note: The article reports 334 complete tracts out of 408, but also reports 112 exclusions. Those quantities are arithmetically incompatible. The included population, exclusion count, and exclusion share must be rebuilt from the source inventory.
Nearly Half the County Is Getting Worse
Let's start with the big picture.
| Trajectory | Tracts | Percentage | Average Change |
|---|---|---|---|
| Deteriorating | 160 | 47.9% | +4.4 points |
| Stable | 107 | 32.0% | +0.1 points |
| Improving | 67 | 20.1% | -3.1 points |
The article reports more than twice as many deteriorating tracts as improving tracts. Because the published tract totals are internally inconsistent, that imbalance cannot yet support a county-wide conclusion.
Something to Note
The article reports that deteriorating tracts went from 5.2% SNAP participation in 2019 to 9.6% in 2023, an 86% increase.
It reports that improving tracts went from 7.6% to 4.5%, a 41% decrease. These rates and their denominator are under reconciliation.
Within the article-reported rates, the improving group starts higher and the deteriorating group starts lower, and their paths cross. That pattern remains under reconciliation.
If reproduced from one tract inventory, the crossing paths would motivate further study. The current values do not establish what changed in those neighborhoods.
Vulnerable Areas Are Falling Behind
We might ask: does baseline vulnerability predict which direction a tract goes?
The article reports a strong association, but the comparison has not been reproduced from a matched output.
| Vulnerability | Deteriorating | Stable | Improving |
|---|---|---|---|
| Q1 (Lowest) | 16% | 67% | 18% |
| Q2 | 35% | 33% | 32% |
| Q3 | 43% | 37% | 20% |
| Q4 | 66% | 15% | 19% |
| Q5 (Highest) | 77% | 11% | 12% |
Two things stand out.
First, the article-reported values suggest a vulnerability gradient: low-vulnerability tracts are mostly stable (67%), while high-vulnerability tracts are mostly deteriorating (77%). This pattern still requires reproduction from a matched output.
Second, the article reports that only 12% of highest-vulnerability tracts improved. If reproduced, that would be consistent with an uneven recovery.
What Do Deteriorating Tracts Look Like?
The article compares demographics across the three trajectory groups. The table values below are article-reported and have not yet been reproduced from a matched output.
| Characteristic | Deteriorating | Stable | Improving |
|---|---|---|---|
| Hispanic % | 28% | 17% | 32% |
| Asian % | 38% | 41% | 33% |
| Bachelor's+ % | 50% | 65% | 48% |
| Poverty rate | 8.5% | 5.4% | 6.8% |
| No vehicle % | 7.1% | 3.8% | 4.8% |
| Single parent % | 41% | 33% | 39% |
Poverty stands out in the article-reported comparison. Deteriorating tracts have 57% higher poverty rates than stable tracts. That association does not establish that economic distress causes worsening SNAP participation.
Education differs too. The article reports 65% college attainment in stable tracts and 50% in deteriorating tracts. That descriptive difference is not a causal result.
Race patterns are mixed in the reported table. Improving tracts have the highest Hispanic share (32%), while deteriorating tracts are slightly below average. The pattern needs reproduction before interpretation.
Vehicle access differs in the reported table. Deteriorating tracts have nearly twice the rate of households without vehicles (7.1% vs 3.8%). This descriptive comparison does not isolate the effect of vehicle access.
Where Is This Happening?
The article reports geographic clustering in specific areas. That pattern has not been verified from a matched public output.
Getting worse:
- East San Jose (Alum Rock, East Foothills, Evergreen)
- South County (Gilroy, Morgan Hill)
- North San Jose industrial areas
Holding steady:
- West Valley (Los Gatos, Saratoga, Campbell)
- Cupertino and western Sunnyvale
- Palo Alto and Los Altos
Getting better:
- Scattered across the county
- Some in central San Jose
- Some in Mountain View near tech employers
The article describes neighboring tracts as following similar paths. That geographic clustering is not publicly reproduced, and shared local causes remain hypotheses.
These Aren't One-Time Spikes
We might wonder: are these sudden changes or gradual trends?
The article reports gradual tract paths over time, but no verified trajectory figure is currently displayed.
Deteriorating tracts: The article reports a steady increase from 5.2% to 9.6%. The unreproduced series and overlapping ACS vintages do not establish sustained worsening or distinguish a pandemic shock from a recovery path.
Stable tracts: The article describes a narrow band. The available materials do not establish that these communities experienced no lasting damage.
Improving tracts: The article describes a higher starting value, a decline from 2019 to 2021, and a later leveling. Economic improvement and household movement are untested explanations.
What's Driving This?
The article reports divergence, but the inconsistent denominator and missing matched output prevent treating it as a reproduced finding. The available tract-level design also cannot identify why the reported paths differ.
Several explanations are possible.
Different Economic Recoveries
High-vulnerability tracts may have more service workers who faced longer unemployment and slower wage recovery.[4] Tech workers returned to high-paying remote jobs. Service workers faced reduced hours and stagnant wages.
This is one hypothesis consistent with the article-reported 15-point college-attainment difference. It is not a tested mechanism.
Housing Squeezing Out Food
As rents rose after COVID, stretched households had to choose between rent and food.[5] SNAP enrollment increased not because incomes fell, but because housing took more of static incomes.
Santa Clara County rents rose 18% from 2019-2023. Deteriorating tracts have higher baseline poverty, suggesting less cushion. Research shows rent-burdened households face roughly three times the odds of food insecurity.[6]
People Moving
Improving tracts may show falling SNAP rates not because residents are better off, but because food-insecure families moved to cheaper areas while higher-income families moved in.
Improving tracts have the highest Hispanic share, and Hispanic households showed net outmigration from the county during this period.
The Honest Answer
We do not know which explanation dominates. Separating them requires household-level data not used here. If the tract-level pattern is reproduced, it still could not distinguish changing circumstances among residents from changes in who lives in each tract.
If the reported tract pattern is reproduced, the affected neighborhoods would warrant closer investigation.
Lessons from These Results
In a county with these patterns, some lessons emerge.
Understanding These Neighborhoods
Understanding food security in deteriorating neighborhoods will depend on understanding SNAP outreach, food pantries, and emergency aid in these areas. The article reports 56 deteriorating Q5 tracts containing about 68,000 residents; those totals inherit the unresolved tract population. Learning more about existing resources and gaps would be a starting point after reconciliation.
Levels and Trajectories Both Matter
Monitoring current SNAP rates is useful. Adding trajectory indicators would provide additional context. A tract at 6% that's been rising 1 point per year tells a different story than a tract at 8% that's been stable.
Study the Improving Tracts
The article reports 67 improving tracts. After the count is reproduced, those tracts could support questions about programs, jobs, or compositional change.
Consider Housing in the Conversation
If housing costs are contributing to food insecurity increases, food programs alone may not be enough. Housing costs and housing-based solutions (rent stabilization, housing vouchers, income support) should be considered in the conversation around food stability.
Limitations
Missing tracts: The published count of 112 exclusions and 27% share conflicts with the reported 408 total and 334 complete tracts. The exclusion population is under reconciliation.
Smoothed data: 5-year ACS estimates average over time. True year-to-year swings may be larger.
Places vs. people: We can't tell whether tract changes reflect residents' changing circumstances or different people moving in and out.
Threshold choice: The 1-point threshold is somewhat arbitrary. Different thresholds would reclassify some tracts.
Summing Up
What We Can't Say Yet
Why this is happening: Is it economic recovery differences, housing pressure, or demographic change? Separating these requires household-level longitudinal data.
Where this is going: Will deteriorating tracts keep worsening, stabilize, or recover? We need 2024-2025 data.
What would help: Would targeted interventions reverse trajectories? Geographic concentration of resources could test this.
What We Can Say
Published divergence claim: 160 tracts deteriorating versus 67 improving, with 77% of the highest-vulnerability group deteriorating. The unresolved denominator prevents a verification claim.
Consistency remains to be tested. A matched rerun must reproduce the threshold sensitivity, vulnerability gradient, and geographic clustering from one tract inventory.
Published scale: The article reports 195,000 residents across 160 deteriorating tracts and a 4.4-point average increase. Those totals inherit the unresolved tract population.
Data: Census ACS 5-Year Estimates 2019-2023. Analysis conducted October 2025.
References
[1] California Department of Social Services, CalFresh Data Dashboard, 2024.
[2] Gundersen, C., & Ziliak, J. P. (2015). Food Insecurity and Health Outcomes. Health Affairs, 34(11), 1830-1839.
[3] Leonard, T., Hughes, A. E., Donegan, C., Santillan, A., & Pruitt, S. L. (2018). Overlapping geographic clusters of food security and health: Where do social determinants and health outcomes converge in the U.S.? SSM - Population Health, 5, 160-170. https://doi.org/10.1016/j.ssmph.2018.06.006
[4] Piacentini, J., Frazis, H., Meyer, P., Schultz, M., & Sveikauskas, L. (2022). The impact of COVID-19 on labor markets and inequality. U.S. Bureau of Labor Statistics Working Paper 551. https://www.bls.gov/osmr/research-papers/2022/ec220060.htm
[5] Denary, W., Fenelon, A., Whittaker, S., Esserman, D., Lipska, K. J., & Keene, D. E. (2023). Rental assistance improves food security and nutrition: An analysis of National Survey Data. Preventive Medicine, 169, 107453. https://doi.org/10.1016/j.ypmed.2023.107453
[6] Brady, P. J., Berry, K. M., Widome, R., Valluri, S., & Laska, M. N. (2024). SNAP Emergency Allotments, Emergency Rent Assistance, Rent Burden, and Housing and Food Security, June 2022-May 2023. Preventing Chronic Disease, 21, 240121. https://doi.org/10.5888/pcd21.240121
Frequently Asked Questions
How many Santa Clara County census tracts are getting worse versus better?
The article reports 160 deteriorating and 67 improving tracts. The percentages and denominator are under reconciliation because the complete and excluded tract counts are inconsistent.
What percentage of high-vulnerability tracts are deteriorating?
The article reports 77% of the highest-vulnerability group and 16% of the lowest-vulnerability group deteriorating. The comparison is not yet reproduced from a matched output.
How much did SNAP rates change in deteriorating tracts?
The article reports that deteriorating tracts went from 5.2% SNAP participation in 2019 to 9.6% in 2023, an 86% increase averaging +4.4 percentage points. These rates and their denominator are under reconciliation.
What distinguishes deteriorating neighborhoods?
The article reports 57% higher poverty rates (8.5% vs 5.4%) and nearly twice the rate of households without vehicles (7.1% vs 3.8%) in deteriorating tracts compared with stable tracts. These comparisons are not yet reproduced from matched output.
Where are deteriorating tracts concentrated?
The article describes possible clusters in East San Jose, South County, and North San Jose industrial areas. That geographic pattern is not yet reproduced from matched output.
How to Cite This Research
Cholette, V. (2025, October 26). The widening gap: Why some neighborhoods are falling behind. Too Early To Say. https://tooearlytosay.com/research/food-security/diverging-trajectories/Copy citation