House Rich, Cash Poor: What 2.7 Million Donations Reveal About Donor Behavior

We set out to prove one thing and found the opposite.

When building our Bene Score™ donor intelligence system, we tested whether demographics could predict charitable giving. The nonprofit industry spends an estimated $500 million annually on wealth screening services built on this premise: find donors in wealthy ZIP codes, and you'll find people with capacity to give.

So we tested it. We analyzed 2,728,689 charitable transactions from 936,377 donors across 711 nonprofit organizations, spanning February 2010 to December 2025. We matched donor 5-digit ZIP codes to U.S. Census Bureau data on home values, income, education, and 14 other demographic variables.

What we found changed everything about how we built Bene Score—and should change how nonprofits think about donor intelligence.

The Wealth Screening Reality Check

Here's what the data showed:

Someone might say: "$100 is better than $53—that's nearly double!"

Fair point. There is a small directional relationship. Wealthier ZIP codes do give somewhat more. But here's why it has almost no practical value:

First, the diminishing returns. You need 4.6x the wealth to get 1.9x the giving. A million-dollar ZIP code yields just $47 more than a $230,000 ZIP code.

Second, the signal is drowned by noise. The correlation (r = 0.004) means that within each ZIP code quartile, you have donors giving $10 and donors giving $10,000. Knowing someone's ZIP code tells you almost nothing about what that individual will do.

Third, the cost-benefit makes no sense. You're paying for wealth screening to identify donors who give $47 more on average. Meanwhile, knowing a donor gave $100 last year—data that's free and already in your CRM—tells you they'll likely give around $100 again. That's 28,000 times more predictive.

The Capacity Illusion

The belief that neighborhood wealth predicts charitable giving. Our analysis of 17 demographic variables found the maximum correlation was r = 0.043 (education), explaining just 0.18% of the variance in giving behavior. Demographics tell you almost nothing.

Why Demographics Fail: House Rich, Cash Poor

Here's the key insight that explains our findings: home value appreciation hits the COST side of household budgets, not the income side.

A $300,000 House in 2010

Mortgage rate: ~4.7% • Monthly payment: ~$1,200 • Property taxes: ~$3,600/year • Total annual cost: ~$20,000

That Same House in 2025 (Now $600,000)

Mortgage rate: ~6.8% • Monthly payment: ~$3,100 • Property taxes: ~$7,200/year • Insurance tripled • Total annual cost: ~$48,000

A wealth screener sees "doubled home value" and concludes "high capacity." But the 2025 buyer has $28,000/year LESS discretionary income than the 2010 buyer—for the exact same house.

The Federal Reserve Data Confirms It

If our analysis were an isolated finding, you might be skeptical. But Federal Reserve data tells the exact same story.

Total U.S. household debt hit $18.59 trillion in Q3 2025. That's an all-time high. Credit card debt alone is $1.23 trillion—also an all-time high, up 60% since 2021.

If home value appreciation were translating into real wealth and capacity, we'd expect to see debt going DOWN as home values go UP. Instead, we see the opposite. Home values at record highs. Credit card debt at record highs. Both at the same time.

This isn't a contradiction. It's confirmation. Rising home values don't create spendable wealth. They create higher property taxes, higher insurance costs, and for new buyers, much higher mortgage payments.

Nearly 4 in 10 homeowners tapping their home equity in 2024 are doing it to pay off other debts. They're using their homes as ATMs to stay financially afloat. A wealth screener sees their $600,000 home and says "high capacity." The homeowner is using that equity to keep the lights on.

What Actually Predicts Giving: Behavior

When we tested behavioral signals—what donors actually DO—the results were dramatically different:

Key Insight

The best way to move forward is to look backward. Your donors' history with you predicts their future better than any external data ever will. And that data is already in your database, for free.

The Relationship Paradox

It would be easy to misread this finding. "Last year's giving predicts this year's"—so just look at the money, right?

Wrong. Those donors gave last year because of the relationship. Their past giving reflects the quality of your past engagement. Without continued nurturing, that predictability degrades.

The finding actually reinforces the importance of relationship building: your best donors need the most attention, not the least.

The 30-Day Conversion Window

Perhaps our most actionable finding: the timing of a donor's second gift determines their trajectory.

We found that 67% of first-time donors never give again—they're one-and-done. (For context, that's actually better than dating: about 80% of dating app first dates don't lead to a second date. Your donors like you more than the average match.)

But look at what happens based on when they make that second gift:

First-time donors who make a second gift within 30 days are 3.3x more likely to become recurring donors. This is the nurturing window—and most organizations miss it entirely. Your first-time donor welcome sequence matters more than any demographic append.

The Engaged Donor Premium

We discovered something else that validated our relationship-first philosophy. When we compared donors who actively choose to give each time (responsive donors) versus those on autopay:

Responsive donors give 1.7x more lifetime value despite giving less frequently. They stay twice as long. Their gift sizes are double. Autopay optimizes for transaction count. Relationships optimize for lifetime value.

What This Means for Your Organization

Rethink your wealth screening investment. Look at what you're paying. Compare it to the results you're achieving. Compare it to the free Census data available at api.census.gov. The data you're looking for is already in your database—your donors' actual behavior. The grass isn't greener elsewhere.

Focus on the 30-day window. Your first-time donor welcome sequence matters more than any demographic append.

Prioritize engagement over autopay. Donors who actively choose to give are worth more—even though they give less often.

Nurture your best donors. The fact that last year's giving predicts this year's isn't permission to ignore them—it's evidence that your relationship worked. Keep working it.

The Bene Score Philosophy

Treat donors as partners, not ATMs. Wealth doesn't predict giving—relationship does. Engaged donors outperform frequent donors. Past behavior reflects past relationship quality. Future behavior requires continued investment.

Methodology & Transparency

We believe in transparency—and falsifiability. Science advances when findings can be challenged, replicated, and improved.

Transaction data: 2,728,689 transactions from 936,377 donors across 711 organizations on the Click & Pledge platform, February 2010–December 2025. Fully anonymized.

Census data: U.S. Census Bureau American Community Survey (ACS) 5-Year Estimates, 2010–2023. 19 files, 373 MB, covering 33,971 ZIP Code Tabulation Areas (ZCTAs). Freely available at api.census.gov/data.html.

Geographic matching: Standard 5-digit ZIP codes. No individual property values or personally identifiable information accessed.

Try It Yourself

Not sure how to download Census data? Ask Claude.ai or any AI assistant:

"Write Python code to download ACS table B25077 (median home value) for all ZIP codes from the Census API."

Working code in seconds. Or do it in Excel—export your donor data with ZIP codes, VLOOKUP the Census data, run a correlation. We'd love to see what you find.

For replication inquiries or questions about this research: Research@ClickandPledge.com

References

1. U.S. Census Bureau. (2010–2023). American Community Survey 5-Year Estimates. Tables B25077, B19013, S1501, S0101, DP04, S1101, S0801. Census API →
2. Federal Reserve Bank of New York. (2025). Quarterly Report on Household Debt and Credit. Center for Microeconomic Data. NY Fed →
3. Click & Pledge. (2010–2025). Anonymized Transaction Data. 2,728,689 transactions, 936,377 donors, 711 organizations.

House Rich, Cash Poor

Hear this research discussed in depth on the Fundraising Command Center Podcast.

Listen to Episode →