A few weeks ago, I drove out to the Coachella Valley—the working-class farm towns east of the festival most people have heard of—because residents there were fighting what would have been the largest data center in California. Initial plans showed a 240-acre+ technology campus, proposed by developer Stronghold Power Systems. The proposed site sat about four minutes from Valle del Sol Elementary School. The City Council had approved the deal months earlier, in a nearly empty room, with no public comment.

By the time the June 4th special meeting that I attended came around, the City Hall was definitely not empty. The chambers were packed, and I and dozens of locals had to watch the meeting from the outdoor courtyard via a large flat-screen monitor they set up for the event. Many of the locals who gave public comment—all of whom opposed it—emphatically mentioned that the development was going to be two miles from the elementary school.

I wondered if this siting situation was unique to Coachella Valley. So afterward, I mapped the location of every American data center project from the nonprofit FracTracker’s data center tracker against the Department of Education’s lists of every public and private K-12 school. It turns out it was not unique to Coachella.

The findings: More than half sit within a mile of a school. Nearly 80% are within two.

The closest in the country is in Atlanta, 121 feet from a school built specifically for children who have experienced homelessness. The next closest hasn’t been built yet: a proposed site in Kansas City, 268 feet from an elementary school (hover over the map for details and sources).

There is no national standard governing how close a data center can be to a school. California maintains a formal regulation—Title 5, Section 14010—dictating how far a school must sit from a high-voltage power line, built on decades of health research. Yet a data center, which can draw as much power as a small city, faces no equivalent rule.

I also visited the Riverside County recorder’s office and pulled the deeds for the Coachella site. The developer paid $7.1 million for the four parcels of land, roughly a third of what data-center land commands nationally. Inexpensive ground, in a town the company appeared to assume wouldn’t push back. (Other media outlets have reported around $17 million, an inaccuracy which may stem from an erroneous calculation counting one deed for three parcels as three distinct prices instead of one price for all three.)

After months of packed council meetings, Coachella killed the deal and passed a moratorium. Organizers—led by local groups like Viva La Causa—are now pushing for a permanent ban, to be discussed by City Council on August 12, 2026.

Coachella isn’t alone. Across the country, data shows the poorest neighborhoods resist data centers at nearly five times the rate of the wealthiest. Similarly, areas with higher rates of homeownership versus renter-heavy neighborhoods were more likely to resist. In fact, the neighborhoods least likely to resist are high-income, renter-dominant neighborhoods— like Manhattan or Downtown Los Angeles, places where land prices tend to be higher and thus more expensive to build on. Which suggests that if communities continue protesting data centers, leading to development delays and cancellations like in Coachella, tech companies may run out of ideal sites to build data centers. (The resistance analysis draws on work by Geoff Holtzman.)

I intend to keep analyzing data and pulling documents—the deeds, the filings, the records few people read—and showing you what’s inside them to trace how land and power affect our lives in California, the U.S., and eventually, across the world.

You can watch the video for this story on my YouTube.

It is already established that the Electoral College isn’t fair on a per-capita basis. As a California resident, I count as one-fourth of a voter compared to someone in the state of, say, Wyoming. (Wyoming has a population of around 530,000 and 3 electoral votes while California around 36 million people and 54 electoral votes out of a total of 538 electoral votes).

Beyond this, the electoral system also carries a fundamental flaw: American taxpayers do not receive fair or equal political representation by state, contrary to the United States's capitalist principles.

Why? It starts with the federal income tax.

By comparing the share of each state’s 2019 federal income tax contributions to its share of Electoral College votes, I found that the electoral system overwhelmingly disenfranchises higher-taxed residents over those paying less. In other words, Americans who live in higher-tax states, like New York or Massachusetts, pay the bulk of federal income tax revenue–but proportionally, their tax dollars tend to earn lesser representation in the Electoral College. (And before you start to generically justify it by saying that those states’ residents are rich and can afford to participate in other ways, may I remind you there are poor folks there too—southern California and New York has some of the worst income inequality in the country—who are also getting cheated.)

Taxation without representation is, after all, the founding injustice that created this country. American colonists, tired of paying taxes to Britain without any say in its political governance, revolted against British rule in the Revolutionary War. Though the colonists won that war and gained independence through the formation of the United States of America, that same political disenfranchisement is in place domestically.

Here is an example to illustrate how convoluted the Electoral College has become: Mississippi is in the lowest tax contribution quartile, yet has among the highest electoral returns per tax dollar. Its 6 electoral votes constitute 1.1% of the map while its income taxes constitute only 0.3% of federal income tax revenue, meaning it receives roughly 3.5 votes to the dollar. 

Mississippians’ tax dollars go much further than New Jerseyians’, who receive only 0.66 votes to the dollar. New Jersey contributes 3.9% of the country’s income revenue while receiving 2.6% of electoral votes.

So for people living in New Jersey, one electoral vote costs $1.50, while for Mississippians it costs only 29 cents.

In 2019, the state with the fairest votes-to-taxes ratio was Delaware (with one vote to one dollar). In the chart and links below, a valuation lower than 1 means its residents pay more in taxes than they receive in fair electoral representation; if it’s greater than 1, they receive more representation than their share of taxes.

[Chart below is here and stats here. Full chart of electoral vote costs here.]

States with the highest electoral returns are West Virginia, Wyoming, Vermont, and Alaska, yet they contribute among the least in taxes. No offense intended, but they’re also coincidentally stewards of receding industries like shale, and neither are they a bellwether of the country’s demographic trends. West Virginia and Vermont, for instance, have some of the highest median-age populations. In fact, the five highest median-age states are overrepresented in electoral returns: Maine, New Hampshire, West Virginia, Florida, and Vermont. 

To be clear, I am not saying older, homogenous populations should have less representation. Today’s reality is the opposite: States with younger populations pay more to be represented. Future generations and growth industries like tech have weaker influence on national politics than those of aging populations and non-critical sectors. Much is made of young people not turning out in high numbers—yet the Electoral College already has their votes diluted.

Democracy should not be pay-to-play. Yet each tax-paying individual’s political representation should be the same as the next person’s, regardless of where he or she lives. The system also must safeguard economic prosperity, since these votes-to-tax returns correlate with votes-to-state GDP returns. Put another way, not only are some taxpayers getting the shorter end of the democratic stick, but so are entire industries.

This means that the current winner-take-all electoral system overemphasizes a few random swing states’ interests at the expense of the rest of the country, a dangerously anti-majoritarian touch. Today, we’re witnessing the disturbing outcome as policies supported by a majority of Americans–regardless of party affiliation–are increasingly overcome by minority rule.

Today, it’s bad enough that I count as only a quarter of a voter from Wyoming in the Electoral College. That my tax dollars, too, are worth five times less than theirs in the presidential election should be a crime. Think of it this way: We get penalized by the IRS if we don’t pay our fair share of taxes, so shouldn’t the federal government also face consequences if they don’t grant our money’s worth of representation in the country’s highest election?

“No taxation without representation” was the original slogan of the United States. Yet it’s an injustice that exists to this day and could stifle our future in the absence of reform. A possible solution is eliminating the Electoral College altogether and operating with a national popular vote. Any others? Let me know.

EDIT 3/2/24: This article was updated with the table chart of electoral college vote costs.

Subscribe now

Leave a comment