Forensic Rebuttal of AI Datacenter Benefits

Investment & Jobs: Heavy Investment, Light Employment

Claim (Proponents): The Kuna data center represents an $800+ million investment, creating around 1,000 construction jobs at peak and ~100 high-paying permanent tech jobs, with multiplier effects of 3–4 indirect jobs per direct job[1][2]. This, they argue, validates Idaho’s incentive strategy and keeps local talent in-state[3][4].

Rebuttal (Reality): Data centers are extraordinarily land-and-capital intensive while generating very few jobs. The Kuna facility will sprawl over 620 acres and nearly 1 million square feet yet permanently employ only about 100 workers – an extremely low employment yield[5]. Even pro-development officials acknowledge these projects are “large land-intensive but not job-intensive” uses that make a city’s economy less efficient[6]. For perspective, that same acreage as a light industrial or business park could host dozens of small manufacturers or offices collectively hiring hundreds if not thousands of Idahoans[7]. Dedicating such vast resources to a mere 100 jobs is a poor trade-off, amounting to roughly one job per 6 acres or $8 million of investment. If Meta were to automate further or pull out, those 100 jobs could vanish, and no other employers are on that site to “pick up the slack” – a huge risk when tying up so much land in a single use[8].

The touted construction jobs are mostly transient and imported – not a lasting local boon. Yes, roughly 1,000 workers swarmed the site during peak construction, but this was a short-lived spike that “largely benefits outside contractors” and traveling specialty crews, not local labor[9]. Idaho’s labor market was essentially at full employment during the build, with virtually no slack to staff a project of this size. Unemployment in the Treasure Valley hovered around 2–3%, meaning almost every additional construction hire had to be an out-of-area recruit[10]. Indeed, Meta brought in a national general contractor (Hensel Phelps) and numerous out-of-state subcontractors for the highly technical work, because Idaho simply didn’t have a thousand idle electricians, pipefitters, and data technicians on the sidelines[9]. One analysis found Idaho’s skilled trades “Slack Index” was effectively 0.0 for electricians – every local electrician was already employed, so “nearly every additional job had to be filled by an out-of-area recruit”[11]. In short, the construction boom did not mean a thousand unemployed Idahoans back to work – it meant an influx of traveling workers who spent a few months in Kuna and then left. By contrast, if that same investment were spread across, say, 50 local business expansions, the construction would likely be phased over time and predominantly serviced by local builders and tradespeople, providing steadier employment for Idahoans rather than a one-time out-of-state blitz[12].

Permanent jobs: quality doesn’t offset the lack of quantity. Proponents stress that ~100 operational roles are “high-paying” and “diversify the local job market”[3], but no matter how you slice it, 100 jobs is a rounding error in a regional labor force. Kuna’s population is about 25,000, the Boise metro is 770,000+ – a hundred jobs hardly moves the needle for broad-based employment. By comparison, a single midsize Idaho manufacturer or a new potato processing plant could match or exceed that job count. Moreover, those 100 data center jobs (mostly technicians and engineers) will require very specific skills; if locals don’t have those skills, Meta will import talent from elsewhere. The claim that this will “keep Idaho’s STEM graduates in-state” is speculative at best[13] – there is no guarantee a Boise State engineering grad will land one of these roles versus an external candidate recruited by Meta. In any case, pinning Idaho’s talent retention on 100 jobs is grasping at straws. The bottom line is that data centers provide a minuscule number of jobs relative to the massive capital invested. As one economic development official put it, these projects “gobble up” land and resources for very little employment, which ultimately makes the economy less efficient[14].

Multiplier effects are overstated and misleading. The article cites an oft-quoted figure that each data center job supports ~3–4 additional jobs in the economy[2], pointing to Loudoun County’s 78,000 indirect jobs from its data center industry. This multiplier comes from an industry-sponsored analysis and should be viewed with skepticism[15]. Many of those “indirect” jobs are temporary construction work or low-wage services (security guards, landscapers, food trucks serving the site) that aren’t long-term, high-quality jobs at all[16]. Any large construction project – a mall, a highway, a new subdivision – will create spin-off jobs; there’s nothing uniquely great about data centers in this regard. Once construction is finished, a data center’s ongoing multiplier is meager because the operation is so automated and staff so few. As independent economists note, the local job ripple from a facility with only 100 employees is negligible – “a mere 100 employees on-site can only generate so much local commerce”[17]. Those workers buying lunch or getting haircuts contribute no more to the economy than 100 workers of any other business. In contrast, a 600-acre business park filled with dozens of employers and thousands of workers would sustain far more secondary jobs (teachers, retail, healthcare) through their spending[18]. This is why studies warn that the job-creation effect of data centers is practically nil relative to their enormous capital cost[19]. The Loudoun County example is an outlier (a massive cluster 20+ years in the making), and even there, many of the “78,000 supported jobs” are likely in sectors that any comparably sized development would generate. Data centers are simply not the labor engines their boosters portray – they are highly efficient for the company (lots of output with few workers) but inefficient for the community looking for widespread employment[20].

Idaho’s costly incentives yielded a poor bang-for-buck on jobs. Proponents herald Idaho’s 2020 tax incentive (sales tax exemption) as crucial to landing Meta’s project[4]. But attracting an $800M investment means little if most of that spend doesn’t recirculate locally. In fact, a huge share of that $800M went to servers, electrical gear, and construction materials purchased from out-of-state vendors, which do nothing for Idaho jobs[21]. Meanwhile, Idaho gave up enormous tax revenue in the deal: the sales tax exemption on ~$500M of equipment/building materials is worth an estimated $30–50 million that Meta did not pay to the state[22][23]. That’s money that would have funded Idaho schools, roads, and public services – forgone to entice one of the richest companies on Earth. And for what? 100 permanent jobs. Idaho essentially wrote a $30M+ check to Meta in exchange for a promise of employment that a single medium-sized local employer could equal[24]. Academic analysis (e.g. Mercatus Center) finds that these kinds of targeted subsidies rarely change a company’s location decision – corporations choose sites for fundamental reasons like power costs, climate, and workforce, not one-time tax perks[25]. Indeed, Idaho has cheap hydropower and cool climate, advantages that likely mattered more to Meta than the tax break. If that’s true, Idaho needlessly gave away tens of millions in public revenue[26]. Even if the incentive did tip the scales, the precedent is set: now every big data center will demand the same deal, a classic race to the bottom where states compete to subsidize companies that produce very few jobs. As Mercatus researchers bluntly put it: “economic development subsidies only help their corporate recipients and the politicians that supply them. Other companies, local residents, and the economy at large are harmed.”[24] In this case, Meta gets a tax-free ride and PR boost; politicians get to claim a win; Idahoans get 100 jobs and a higher bill for infrastructure (as discussed later). Viewed objectively, the cost per job of this deal is astronomical – hardly the triumphant “strategic boon” advertised.

Infrastructure: Upgrades or Burdens?

Claim (Proponents): Hyperscale data centers spur major infrastructure investments that benefit the whole community. In Kuna, Meta paid ~$70 million for a new water/wastewater treatment plant (handing it over to the city) and implemented water reuse systems, expanding local utility capacity[27][28]. Data center projects also drive improvements to the electric grid (e.g. a new substation, new renewable energy projects) that improve reliability for residents[29][30]. Meta’s commitment to 100% renewable energy led to a 200 MW solar farm in Idaho and a new green tariff program, ostensibly boosting Idaho’s clean energy portfolio for other users as well[31]. In short, the community gets new infrastructure and greener power largely paid for by the company’s investment.

Rebuttal: The “free” infrastructure isn’t free at all – it’s mitigation for the data center’s immense demands, and other upgrades will ultimately come out of Idahoans’ pockets. It’s true Meta paid to build a water treatment facility and related infrastructure in Kuna. However, this wasn’t charity or excess capacity out of goodness; Meta needed that plant to supply its own operations. The east Kuna sewer upgrade was fast-tracked because the data center couldn’t function without it. Yes, the city now owns a modern utility asset, but calling it a community “benefit” is disingenuous when it was a prerequisite for Meta’s project[32]. As Idaho’s data center initiative report notes, Meta’s ~$50M investment in Kuna’s water/sewer “covers only the data center’s immediate needs”, enabling their campus and an adjacent industrial park – it does not offset the broader strain on the overdrawn aquifer or existing city systems[32]. In essence, Meta paid for its hookup – a cost any large industrial user would bear. Framing this as a gift to the public is like applauding a factory for paving the access road to its facility; it’s standard mitigation, not a net positive for unrelated residents.

The same goes for the power infrastructure. The pro-AI article highlights a new substation and a 200 MW solar farm tied to the data centers[33]. Again, these investments are driven by the project’s own requirements. Meta’s 100% renewable pledge meant it procured a dedicated solar farm – but that solar array will only cover a fraction of the data center’s load[34]. A single hyperscale AI data center can draw on the order of 100+ MW of power continuously (enough for tens of thousands of homes)[35]. A 200 MW solar farm doesn’t produce 200 MW at night or on cloudy days, and its output feeds generally into the grid, not exclusively to Kuna. Idaho Power’s own planning documents reveal that the influx of large data centers is driving a projected 20% increase in statewide electricity demand within a decade, a surge largely attributable to these new facilities[34]. The utility simply does not have that much spare capacity lying around. This means expensive new generation resources and grid expansions are needed. In fact, Idaho Power has outlined a $4 billion grid upgrade plan (new power plants, transmission lines, etc.) predominantly to meet data center and other high-tech loads coming online[36]. Who will pay for those billions? In regulated utility finance, customers do. The utility will invest in infrastructure and then recoup costs plus profit through electric rates. Meta will pay its share through special contract demand charges, but if any costs can’t be isolated, or if Meta’s load ever fluctuates, other ratepayers could end up footing the bill[37][38].

Even with “hold harmless” clauses, there’s risk of cost-shifting. Advocates claim contracts will ensure other customers don’t subsidize the data center’s power needs (“no cost to existing ratepayers”). It’s true the Idaho PUC required protective provisions – Meta must guarantee certain payments, cover new transmission build-outs, etc., to theoretically shield regular customers. But the real world doesn’t always stick to theory. Idaho has learned this the hard way: when a large industrial project goes belly-up or underperforms, the utility’s sunk infrastructure costs have, in past cases, been passed to the public. A cautionary example is the Hoku plant in Pocatello – the utility built a substation and upgrades expecting a large load, Hoku even paid some upfront, but then the company went bankrupt. Result: the utility was left holding unused capacity and sought to recover costs from other customers[38]. Despite Hoku’s contractual obligations and posted security, “some costs ended up being socialized” when the project collapsed[39]. While Meta is a far more stable company, the principle remains: if Meta or future data center tenants ever exit early, downsize, or negotiate lower rates later, Idaho Power’s investments could become stranded assets that everyone else has to pay for. Even short of a worst-case scenario, there are subtler burdens – for instance, the wear-and-tear on shared transmission lines or the accelerated need for new energy resources, which are hard to pin solely on one customer in rate cases. Without a strong independent consumer advocate in Idaho to scrutinize these deals, the public interest may not be fully protected. Idaho is one of only three states with no dedicated consumer counsel for utility matters – no agency that solely represents ratepayers at the PUC[40]. That means these complex “hold harmless” arrangements get far less adversarial examination than in most states. The Commission’s staff does its best, but there is no independent watchdog exclusively fighting for residential and small business customers[41][42]. In such a lax oversight environment, one has to trust the utility and Meta to get everything right – a risky bet when billions in infrastructure and long-term monopoly profits are on the table. The bottom line is that any major infrastructure expansions spurred by the data centers will ultimately rely on the public’s financial support, either through utility rates or local taxes, if not immediately then over time. “Upgrades” are not benevolent gifts when they are built to serve the project’s own massive appetite for resources.

Tax Revenue: Overstated Gains and Hidden Costs

Claim (Proponents): The data center greatly expands the property tax base, yielding millions annually for local government and potentially even lowering homeowners’ taxes. Kuna’s data center is cited as adding ~$725 million in new taxable value, enough that the “average homeowner’s property tax bill could drop by ~4.8%” once it’s fully on the rolls[43]. Officials pointed to an estimated ~$110 annual tax reduction on a $535k home – tangible relief credited to Meta’s investment[44]. Moreover, unlike housing developments, data centers “place relatively light demands on services” (no schoolchildren, etc.), making them fiscally attractive[45]. State legislators and city leaders touted this as a “net win”: a high-taxpaying facility that doesn’t strain public services[45]. In addition, companies often sweeten deals with direct community payments – e.g. Diode’s proposed tech park included $40 million to Kuna’s schools, police, and fire services over 20 years[46], and Meta has given $500k to local nonprofits and schools since 2022[47]. These are windfalls that wouldn’t exist without the data center projects.

Rebuttal: Yes, the data center will pay property taxes – but the benefit to locals is far smaller than advertised, and it came after maneuvering that nearly siphoned those taxes elsewhere. The oft-repeated “4.8% tax drop” for homeowners sounds great in a vacuum. In reality, that figure amounted to only about $110 per year for the average homeowner[48] – a modest rebate, not a life-changing windfall. Kuna residents initially had much grander expectations (some hoped for double-digit tax relief); the actual impact fell “far below” those hopes, as Kuna’s own economic development director acknowledged[48]. Why so low? Because even a $725 million project, as huge as it is, gets diluted across the entire tax base and constrained by Idaho’s property tax laws. And that 4.8% reduction was only assured because state lawmakers intervened to force Meta’s full value onto the normal tax rolls[49][50]. Behind the scenes, Kuna had initially put Meta’s site into an Urban Renewal District (URD) – a mechanism that would have diverted the new tax revenue to infrastructure projects rather than to general tax relief[51]. In other words, absent legislative outcry, homeowners might have seen no tax reduction at all; the city was planning to reinvest Meta’s taxes into roads and sewers for an industrial park. Idaho legislators felt this violated the “deal” underlying the 2020 tax break law (which assumed data centers’ value would help the general tax base)[52]. The result was a new state law (HB 328) forcing Kuna to remove Meta from the URD[50]. So the touted property tax relief didn’t happen organically – it had to be politically engineered, effectively overriding local planning. This tug-of-war underscores that the tax benefits are not as straightforward as the headlines suggest. The city lost the chance to use that tax increment to attract a cluster of businesses (which might have ultimately grown the tax base more), in exchange for giving current residents a relatively small tax break now[53]. It’s a policy choice that can be debated, but the key point is: even with the data center on the tax rolls, the benefit to each taxpayer is modest, and significant potential long-term gains were sacrificed to deliver that modest benefit.

Furthermore, big industrial taxpayers often don’t stay cash cows forever. Data centers depreciate rapidly – the servers and equipment decline in value each year, and many of those assets aren’t even taxed (since Idaho exempted them from sales tax and they’re often expensed). The building shell might also get reappraised downward if the facility underperforms or ages. Idaho’s property tax system and politics add more uncertainty: large companies routinely appeal their assessments to lower their tax bills, and Idaho has caps and forgiveness mechanisms that can limit how much of that $725M actually gets taxed at full value[54]. In some states, data centers have secured special valuation discounts or abatements over time; there’s nothing stopping Meta or others from lobbying for future tax relief if economic conditions change. So the projection of X million per year in taxes assumes best-case, steady-state operation. If Meta ever idles part of the campus or negotiates a deal, those tax dollars could shrink. Meanwhile, the community incurs ongoing costs directly tied to the data center: heavy trucks during construction have chewed up local roads, first responders need training/equipment for potential data center emergencies (electrical fires, hazardous materials from batteries or cooling systems), and local governments must stand ready to serve a massive facility even if it employs few people[55]. These costs – wear and tear, specialized emergency readiness, etc. – are real, yet they aren’t line-itemed and reimbursed by Meta’s property tax checks. When boosters say the data center doesn’t burden services because it has no schoolkids, they conveniently ignore those unique burdens it does create (e.g. a 960,000 sq ft high-security site is not your average commercial property for the fire department).

The $40 million “community contributions” are not charity – they’re essentially impact fees spread out over decades. It’s true that Diode (the developer behind another data center park proposal) pledged $40M to local services in its development agreement, and Meta has made some voluntary donations. But let’s unpack that: Diode’s $40M is to be paid over 20 years[56]. That averages $2M a year, of which $1.5M/yr is earmarked for the fire district[57]. Why so much to fire? Because a huge, complex industrial campus is being plopped in a rural area that previously had limited fire/EMS capability. The developer is essentially paying the salary and equipment for additional firefighters to protect its own facility. It’s not a “bonus” to the community – it’s the price of making the project viable. As the rebuttal analysis pointedly asks, is that $30M to the fire department really a “community benefit” or just mitigation of a risk the project imposes?[58] Similarly, the police funding in that package would likely cover one or two extra officers to patrol the remote site and its vicinity[59]. These contributions are better viewed as negotiated impact fees: the company pays to offset the strain its project will create on public services. This is standard practice for large developments (cities often require new subdivisions to pay for new roads, schools, etc.). What’s unusual is that data center boosters tout these payments as if they were altruistic gifts. In reality, “it’s a pay-to-play model” – the company pays the city so that officials will approve the rezone and give the green light[60]. Once the project is built, that money is spent, and the long-term tax revenue from a data center still pales in comparison to, say, dozens of smaller businesses that each pay property taxes and continuously patronize local suppliers. Crucially, small businesses don’t get 20-year deals for special treatment – they pay their full freight of taxes and often have to fund their own growth improvements, whereas these tech giants negotiate custom arrangements and still get tax breaks on top[61].

Statewide, the data center tax incentive undermines public revenues. As discussed earlier, Idaho waived an estimated $30–50 million in sales taxes to woo Meta[22]. That is a direct hit to the state treasury – effectively a subsidy paid by all Idahoans (in the form of forgone revenue) to benefit Meta’s project. If another big data center comes, they’ll ask for the same deal, potentially draining even more funds that could have gone to education or tax relief for citizens. To its credit, Idaho did amend the incentive in 2025 to impose a sunset (new projects only get 7 years tax exemption)[62]. But that came after Meta was already grandfathered in for a perpetual exemption[63]. So for decades to come, Meta’s expansions will avoid sales tax, meaning Idaho will continuously lose out on revenue as the facility upgrades hardware. When politicians claim a project will “pay for itself” in tax base, remember that Idaho’s deal also ensured the project doesn’t pay millions in other taxes. When you tally it up, the net public benefit is murky – a small break for Kuna homeowners, some local fees paid to fire/police (to cover new burdens), but a large break given to Meta at the state level. What is clear is that the opportunity cost was significant: had that same site been developed into a business park or kept for higher-employment uses, the aggregate property tax and economic activity might well exceed what one data center will ever deliver. Instead of many businesses gradually adding to the tax base, Kuna opted for one big player. If that player’s contributions don’t live up to the hype (or if they find ways to minimize their tax valuation), the community has bet big and could lose big. In sum, the tax revenue story is far more complicated than “big data center = big tax win.” After all the special deals and offsets, locals are left with only a modest benefit, while the long-term risks and missed opportunities are downplayed.

Community Impact & Quality of Life: A Tech Giant in the Neighborhood

Claim (Proponents): Meta and other data center companies integrate into the community and enhance local quality of life. They point to Meta’s community grants and philanthropy – over $500,000 donated to Ada County schools, nonprofits, STEM education, and community groups since 2022[64][65]. Meta has launched programs like Community Action Grants in Kuna to fund technology and education initiatives, partnered with the Idaho STEM Action Center, and sponsored local events[65][66]. Company employees volunteer locally, and Meta joined area chambers of commerce, signaling they want to be a true “community partner” not an isolated corporate actor[67]. City leaders say Meta’s presence put Kuna “on the map” for other high-tech investment and even instilled civic pride – being home to a cutting-edge AI data center makes the town special[68]. From a quality-of-life perspective, supporters argue that data centers are “among the least disruptive” industrial uses – they don’t emit pollution, generate little traffic once built, and are quiet neighbors (certainly preferable to, say, a noisy factory or a dense housing subdivision)[69]. At public hearings, some residents favored the data center over alternatives for those reasons[70]. The idea is that Kuna gets the economic benefits without many of the lifestyle downsides other growth could bring.

Rebuttal: The feel-good community narrative doesn’t hold up to scrutiny – the data center’s local philanthropy is token at best, and its presence does little to enrich daily community life. Meta’s much-publicized charitable giving in the area amounted to $500k over a couple of years[47]. For context, that is just 0.06% of Meta’s $800 million capital spend on the Kuna facility[71]. In other words, a drop in the bucket – essentially a PR budget line item. By comparison, Idaho’s existing small businesses collectively donate far more to their communities every year, and they do it out of genuine local commitment, not as a calculated goodwill strategy. According to the U.S. Chamber of Commerce, 75% of small businesses donate to local charities (averaging 6% of their profits) and, proportionally, “small businesses donate 250% more to local nonprofits and community causes than larger businesses”[72]. This makes sense: small business owners live in town, coach the little league, sponsor the county fair, and so on. They are embedded in the community’s social fabric. Meta, by contrast, is an out-of-town behemoth. Writing a check to the Kuna School District or funding a STEM camp is nice – and surely appreciated – but it’s an arm’s-length form of engagement[73]. It doesn’t replace having home-grown companies whose owners and employees are personally present at PTA meetings or city council, deeply invested in the community’s long-term well-being[73]. Meta’s grants are essentially a corporate relations effort. Even a half-million dollars in school grants, spread over multiple schools and projects, is relatively small – one new school gym or a handful of teacher salaries can eat that up. Meanwhile, the 200,000+ small businesses in Idaho each contribute in their own sustained ways, building the community from the ground up. A single corporate outpost just can’t match that organic involvement.

Rather than integrating into local life, the data center risks turning Kuna into “Anywhere, USA.” Community advocates often warn that when a town becomes dominated by outside-owned facilities or chain businesses, it loses its unique character[74]. We can already see this with the data center: it’s a self-contained fortress with security gates, run by a company headquartered thousands of miles away. It’s not going to have a friendly local owner hosting block parties. In the words of one analysis, a community filled with these remote facilities becomes “Anywhere, USA” – a place lacking unique local character, where one town feels the same as the next because the economy is homogenized by national chains or single-purpose facilities.”[75] If Kuna’s future development hinges on more data centers (as officials hope for a broader “Tech Park”), that could crowd out the diverse local entrepreneurs who actually give a town its soul. There’s no Main Street bustle around a data center – it sits there behind berms and fences, contributing almost nothing to the daily social or cultural life. You can’t walk into it, you can’t shop or dine there, you likely won’t work there unless you’re one of the few tech employees or security guards. As one stark summary put it, a data center is a “black box” in the community – “few locals work there, and its contribution to daily community life is essentially nil (you can’t shop there, eat there, or even enter it without security clearance)”[76]. It might as well be a giant warehouse or server farm in the middle of a field – which is literally what it is. This is a far cry from the picture of a vibrant community partner that the company likes to paint.

Civic pride doesn’t pay the bills, and being a “tech hub” of server farms offers little to residents. It’s understandable that Kuna’s leaders are excited to attract a big-name company like Meta. Certainly it put Kuna in some news articles and maybe drew interest from other developers. But we have to separate the symbolic prestige from the practical reality. Yes, not many towns can say they host part of the backbone of the AI-driven internet. But how does that benefit the average resident of Kuna or Idaho? It doesn’t create a broad base of jobs; it doesn’t spawn local startups (since the tech is all internal to Meta’s operations); it doesn’t even provide services to the local population (the data center isn’t processing Idaho data – it’s serving Meta’s global business). The “follow-on” interest we’re seeing – like the Diode Gemstone project – is just more of the same: outside firms wanting to build more data centers on Idaho land. That doubles down on the one-note industry rather than diversifying the local economy. Contrast this with a scenario where Kuna attracted, say, a cluster of food processing companies or an aerospace manufacturer and its suppliers – industries that create ecosystems of varied employment, skills transfer, and local business-to-business commerce. With data centers, the ecosystem is thin. You don’t get a bunch of supporting local companies popping up to feed off the data center; the supply chain (servers, software, cloud services) is all global and internal to Meta. At most you get some local janitorial contracts or landscaping gigs. That’s not how you grow a robust local economy. So the fact that Kuna is now “on the map” to other data center developers is a dubious win – it could lock the area into a path of low-employment uses, making it harder for other industries to take root on that land.

Quiet, low-traffic, “clean” – true, but essentially because the facility is economically inert. Proponents laud that a data center doesn’t clog roads or spew pollution. It’s worth noting the irony: those positives are a direct consequence of the lack of people and production. A ghost town is quiet and has no traffic, too. That doesn’t mean you’d base your economy on it. If minimal activity is the selling point, it underscores how little the project contributes. Yes, neighbors might prefer a data center over a massive subdivision if roads and schools are currently inadequate[70]. But that’s really an indictment of planning – it reflects that local infrastructure hasn’t kept pace with growth, so residents feel they must choose the lesser of two evils. It doesn’t mean a data center is the best use of land; it means officials need to improve roads and schools if they ever want higher-impact development. By acquiescing to a low-impact facility because “our roads can’t handle housing,” Kuna is essentially capitulating on infrastructure improvements that could benefit residents in the long run. Also, while data centers are relatively quiet, they aren’t silent – they have large industrial cooling equipment and backup generators. Neighbors might hear low hums or have concerns when dozens of diesel generators fire up for testing or power outages. It’s not a smokestack industry, but it’s not completely without environmental footprint (diesel particulates from backups, heat island effects, etc.). More importantly, no traffic = no employees = no local commerce. A factory with 500 workers does create traffic – and those workers also patronize local stores, enroll their kids in local schools (which brings state education funding), and generally contribute to the town’s vibrancy. A data center’s ideal state is to hum away unseen, which from a lifestyle perspective might be benign, but from an economic perspective is almost ghost-like. In essence, Kuna traded potential hustle-and-bustle (with its attendant challenges, sure) for a big, fenced-off quiet zone. That’s a valid choice for those who prioritize a sleepy status quo, but it’s not the recipe for broadly shared prosperity or a dynamic community. As one Phoenix official noted when rethinking data center proliferation: these facilities risk hollowing out the economic potential of valuable industrial land[14]. You end up with a lot of land tied up in use that doesn’t engage the community, doesn’t employ many people, and doesn’t synergize with other local businesses. That might be “quiet progress” to some, but it’s also a path to a very one-dimensional local economy.

Sustainability & Resource Impacts: Green Promises vs. Resource Realities

Claim (Proponents): Meta and similar data centers are committed to sustainability and will have minimal environmental impact on Idaho. Meta touts that its operations are powered with 100% renewable energy (net-zero carbon) and it actively invests in new green energy projects in Idaho, like the 200 MW solar farm, which helps advance Idaho’s clean energy goals[77]. On water, Meta emphasizes on-site conservation – using advanced water reuse and recycling for cooling – and has pledged to be “water positive” by 2030, meaning it will restore more water than it consumes[78][35]. The Kuna data center includes innovative wastewater recycling (treated effluent used for irrigation) to reduce freshwater draw[79]. Supporters also note that data centers have no smokestack emissions and, once built, relatively low environmental footprint compared to traditional heavy industries or huge residential developments. In sum, they frame the data center as a clean, high-tech industry that aligns with a sustainable future, turning what might seem like a resource concern (big power user, big water user) into a net positive via green investments and technology.

Rebuttal: The “green” narrative is heavily spun – the data center will actually be one of the largest new strains on Idaho’s energy and water resources, and the much-advertised renewable and conservation measures only partially mitigate that. Start with electricity: Meta’s pledge of 100% renewable energy is achieved on paper by funding equivalent renewable generation, but the facility itself will draw a huge amount of power from Idaho’s grid, 24/7. The 200 MW solar farm Meta enabled is significant – it increases Idaho’s solar capacity – but solar output fluctuates and generally covers daytime hours. Meanwhile, a single hyperscale AI data center can easily require on the order of 100–150 MW of constant power load[35]. At full tilt, Meta’s center could be consuming as much electricity as all the homes in a mid-sized Idaho city. When the sun isn’t shining or the solar output is low, Idaho Power still has to supply Meta with reliable energy, which often means drawing from other sources (including importing power from out-of-state, some of which is not renewable). In other words, Meta’s “clean energy” claim doesn’t mean the data center runs directly on solar 24/7; it means Meta buys enough green credits or output to offset its usage annually. The actual electrons powering those servers at 2 AM or on a cloudy winter day are likely coming from the broader mix of the grid. Idaho Power’s own planning documents explicitly warn that new data centers will drive a 20% increase in Idaho’s electricity demand over the next decade[34]. That’s enormous – roughly equivalent to adding hundreds of thousands of households worth of consumption. It will force Idaho to ramp up energy development at an unprecedented pace, including possibly natural gas plants or expensive imports, if renewables and storage can’t be built fast enough. Proponents cite the new “Clean Energy Your Way” tariff as a benefit (Meta helped create a green power option for big customers)[80]. But note: that tariff is a special program – it doesn’t necessarily lower costs for regular folks, and if anything, it could allow other large industrials to also demand dedicated green power, further straining Idaho Power’s resource acquisitions.

Most critically, the burden of ensuring reliable power falls on the utility and, by extension, all its customers. Data centers demand ultra-high reliability (down to milliseconds of outage tolerance). Idaho Power will likely invest in grid upgrades – new transmission lines, substations (like the one proposed for Gemstone), voltage regulation equipment, etc. – to meet those standards. While Meta and others might pay for some dedicated facilities, the interconnected nature of the grid means a lot of these improvements benefit the data center primarily while being socialized over time. We already discussed the planned $4 billion in system upgrades largely driven by data center growth[81]. Even if special contracts make the companies pay upfront, those assets go into the rate base that all customers fund long-term. Idaho doesn’t even have renewable energy mandates or a robust climate policy driving such investment – it’s being driven by the private demands of these tech firms. Ironically, Idaho residents could end up subsidizing infrastructure that chiefly serves corporate data hubs exporting services worldwide. And while Meta’s solar project is a plus, it only covers a fraction of the data center’s total energy consumption[34] (to say nothing of the multiple other data centers Idaho is courting). So the net effect could be an increase in Idaho’s carbon footprint or reliance on out-of-state power in the near term, even as Meta pats itself on the back for renewables.

On the water side, the discrepancy between PR and reality is stark. Meta publicized that the Kuna data center would “average” 60,000–70,000 gallons of water usage per day[82], thanks to recycling and efficient cooling designs. What they didn’t advertise is that during peak summer conditions, cooling a nearly million-square-foot server farm could require over 1 million gallons per day[82]. Independent analysis of the facility’s cooling needs found potential peak water withdrawals on the order of 1.1 million GPD[83][82]. To put that in perspective, that’s as much water as about 10,000 Idaho residents use in a day (100 gallons per person). This water is drawn from the local aquifer – the Snake River Aquifer – which is already overtaxed. Idaho’s aquifers are under serious strain, with the Snake Plain aquifer overdrafted by roughly 250,000 acre-feet annually (that’s billions of gallons per year being depleted)[84]. Adding a large, continuous industrial water user makes the problem worse, unless offset by significant mitigation. Meta’s promise to be “water positive” by 2030 – meaning they will fund water restoration projects – rings hollow for Idaho specifically. As of late 2022, Meta had provided zero details of any Idaho-based water restoration; they pointed to projects in Arizona, New Mexico, and Texas instead[35]. So Idaho’s water table takes the hit now, on faith that Meta will eventually do something, somewhere to put water back into nature. Even if they do, those out-of-state projects don’t replenish the Boise-area aquifer that Kuna relies on. In short, the data center will consume huge amounts of local water, and the much-touted “water reuse” measures (like using treated effluent for irrigation) only partially reduce the thirst. At the end of the day, a million-gallon-a-day user is a million-gallon-a-day user – that water comes out of Idaho’s allocation, potentially affecting farmers and other users, especially in drought conditions. By comparison, if that land were a mix of small industries or homes, the total water usage might actually be less (farms use water too, but there are regulatory frameworks for irrigation rights, whereas industrial groundwater pumping for cooling is a newer challenge Idaho hasn’t fully grappled with in public).

The mitigation measures (solar farms, water plants) are responses to the project’s impacts, not net-new benefits. It’s important to frame these correctly. When Meta builds a solar farm or a wastewater recycling system, those are requirements* to make a very resource-intensive project palatable. It’s not like Idaho wanted a solar farm and Meta kindly obliged out of community spirit – Meta needed renewable energy to claim green status, so they paid for the solar farm as a cost of doing business. The benefit to Idaho is incidental. Similarly, Kuna’s new water treatment facility was necessary because without it the city’s existing system could not have supplied the data center’s needs[27][32]. Meta didn’t gift a random infrastructure upgrade; they paid for their hookup and then transferred it to the city (which will now incur maintenance costs, by the way, paid over time by local ratepayers). The pro-data center narrative spins these as pluses, but they are fundamentally mitigation – the company covering the direct impacts of its project**. This is akin to a factory installing a filtration system so it doesn’t pollute the river; you wouldn’t normally count the filter as a community “benefit,” it’s just preventing harm. Yet with data centers, because the harms (huge power draw, huge water draw) are somewhat abstract or off-site, the mitigations (new power infrastructure, new water infrastructure) get touted as public enhancements. It’s a clever reframing, but don’t lose sight of causality: if Meta weren’t consuming 100 MW of power, Idaho Power wouldn’t need to build new solar farms and transmission lines; if Meta weren’t using a million gallons of water, Kuna wouldn’t need a new treatment plant and wells.

Finally, consider the broader resource trade-offs. Idaho’s economy historically rests on agriculture and other value-added industries that themselves need water and power. Allocating finite resources to data centers means those resources aren’t available for other uses. For example, that electricity could have gone to electrify transportation or attract a manufacturing employer – uses that create more jobs per megawatt. The water being evaporated to cool servers could have irrigated crops that support Idaho’s $20 billion ag sector. In fact, agriculture contributes on the order of 12–17% of Idaho’s GDP and tens of thousands of jobs[85]. Data center advocates have essentially said: let’s divert some of that water and power to a handful of server warehouses because it’s “high tech.” From a statewide perspective, is that wise? One report pointedly asks whether pouring finite water/energy into data centers is truly the “highest and best use” of those resources, given the opportunity cost to an economy built on other industries[85]. The answer is far from clear. If Idaho experiences power shortages or has to buy expensive emergency power in a hot summer, or if groundwater levels continue dropping, people will rightly question why we allocated so much to servers that don’t even employ many Idahoans.

In summary, the environmental and sustainability claims around Idaho’s AI data centers are more spin than substance. The data centers will materially increase the load on Idaho’s grid (necessitating new energy development that all customers fund) and draw heavily on water supplies (with no local replacement assured). The much-promoted green measures – solar farms, water recycling – are partial offsets required to make the project acceptable, and they often only cover a fraction of the actual resource consumption. Data centers may be quieter and cleaner than a coal plant, but they are far from impact-free. And unlike a truly local enterprise, they extract resources to deliver services and profits that flow mostly out of state. When Meta’s proponents boast of net-zero emissions and water positivity, those refer to global corporate goals with accounting tricks, not to the concrete reality in Idaho. Locally, Idahoans will bear the resource burdens (and risks, if anything goes awry) while getting a scant portion of the economic rewards. A shiny “green” data center can thus become a resource drain on a region, all while the company gets to burnish its sustainability image. Idaho should weigh those facts heavily against the thin benefits when evaluating whether hyperscale AI data centers are truly a boon or just a new breed of extractive industry in digital guise.

[1] [2] [3] [4] [5] [6] [7] [8] [9] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [24] [25] [26] [27] [28] [29] [30] [31] [33] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [79] [80] Idaho’s AI Data Centers_ Promises and Realities.docx

file://file-SaX5g5xiGvpPvSx1bwctQn

[10] [11] Meta Kuna Data Center_ Imported Construction Labor and Local Labor Scarcity.pdf

file://file-NSfRc1py3g51HHBNmv44aE

[23] [63] Set-Up_ What’s at Stake in Idaho.pdf

file://file-AcyrJWPgX7PNuAPxnwZKfB

[32] [34] [35] [36] [78] [81] [82] [83] [84] [85] Idaho’s Data Center Initiative_ Risks, Resource Constraints, and Hidden Costs.docx

file://file-4DwQB7NSV3sP4aGn54hFiJ

[37] [38] [39] [40] [41] [42] Serving AI Data Centers in Kuna_ Idaho Power’s Plan and Ratepayer Impacts.pdf

file://file-Cuyo3yaVeC5aSv83tMQBbJ

[62] Meta’s 1 Million Sq Ft AI Data Center in Idaho – Power, Costs, and Tax Incentives.docx

file://file-8edSP3PesWzeqvrACdco9p