How Data Centers Affect Local Communities

What does it mean for a local economy when a multi-billion dollar data center is set up in a region. Mississippi is currently debating the impact of an upcoming $10 billion Amazon data center. The reality is nuanced and the $10B figure can mislead one into believing a massive job boom is coming. Let's look into it: 1. Direct employment: Impact is usually minimal. From the stats I have seen, ~100 jobs created from new data centers. E.g., Apple's $1 billion data center in Maiden, NC, supposedly created 50 direct jobs (see comments). $10B data center will mean many more direct jobs but unlikely to be in the thousands. Most of the investment goes into hardware and utilities as opposed to direct service jobs. 2. Temp jobs: No doubt there will be many temp jobs including in construction. 3. Tax revenue: This is probably the most significant benefit: taxes on property, construction, etc. The big question is how much tax rebates are being offered to bring the data center to your region. 4. Spillover effect: A high-speed data center could potentially attract high-tech industry and create software jobs. I looked at a heat map of data centers and very few of these places are big software hubs. So it's not a given. If a local economy hopes data centers will create a job boom, there really needs to be a strategy for how it'll be used to attract tech industry, tech talent, etc. Otherwise, net job impact will be low. Northern Virginia is a good case study of both the spillover effect and healthy tax revenues. 5. Energy costs: Data centers are energy hungry and will drive up energy costs for locals. One report says 70% increase in electricity costs (see comments) but I think the numbers are likely much lower. {While there will undoubtedly be a huge climate impact, the biggest long-term positive impact one can hope for is that data centers become the catalyst for safe nuclear energy in the long run.} We are bound to see lots of stories about new data centers being set up all over the world. Job creation for the local economy will be how it's marketed. The real story of economic benefits to locals is rather nuanced though. I shared some comments on this for this story from an NPR affiliate: https://lnkd.in/e-Wy2esP

Between Food Security and Data Centers: Can We Afford to Choose? In Pontian, Johor, a pineapple farmer named Ali Maju tends to his crops using water from canals and underground sources—like he has for over a decade. Just 5km away, massive data centers rise from the ground. Silent, windowless, and powerful, they store everything from AI outputs to your online shopping data. But to function, they must stay cool—requiring millions of liters of clean water every single day. The scale is staggering: As of 2024, Malaysia received 101 applications for data centers, which together require 808 million liters of water daily. Only 45 were approved, but they still demand 142 million liters/day—enough for over 500,000 people. And with droughts expected as early as 2025–2026, the battle over water has begun. But what about the economic payoff? We were promised 41,000 digital jobs. Yet experts warn: actual job creation is minimal. In Sweden, Facebook promised 30,000 jobs, only hired 56. Many data centers are automated, offering short-term construction work, but limited long-term employment. So what are we sacrificing? ➡️ Our food security, as farms like Ali Maju's rely on the same water. ➡️ Our truth in policy, when job numbers are inflated. ➡️ Our sustainability, if energy- and water-hungry infrastructure expands unchecked. And now, a bigger question looms: When drought strikes—who gets the water? In July 2024, YTL Power acquired Ranhill Utilities, the company managing all 46 water supply plants in Johor. This raises a critical concern. Will data centers get priority? Or homes, farms, and everyday people like Ali Maju? Malaysia's digital ambition must not come at the cost of its citizens’ basic needs.

𝗛𝗼𝘄 𝗗𝗮𝘁𝗮 𝗖𝗲𝗻𝘁𝗲𝗿𝘀 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺 𝗟𝗼𝗰𝗮𝗹 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝘁𝗶𝗲𝘀 When a data center moves into a community, the benefits extend far beyond technology: 𝗗𝗿𝗶𝘃𝗶𝗻𝗴 𝗚𝗿𝗲𝗲𝗻 𝗘𝗻𝗲𝗿𝗴𝘆 𝗧𝗿𝗮𝗻𝘀𝗶𝘁𝗶𝗼𝗻: 🎯 Data centers increasingly rely on renewable energy sources like wind, solar, and nuclear. 🎯 In 2023, 81% of new utility-scale renewable projects were cheaper than fossil fuels, lowering energy costs for everyone. 𝗕𝗼𝗼𝘀𝘁𝗶𝗻𝗴 𝗟𝗼𝗰𝗮𝗹 𝗘𝗺𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁: 🎯 In Virginia alone, data centers supported 78,000 jobs and generated $31.4 billion in economic output in 2023. 🎯 Providers actively retrain military veterans and partner with universities for specialized education. 𝗔𝘁𝘁𝗿𝗮𝗰𝘁𝗶𝗻𝗴 𝗔𝗱𝗱𝗶𝘁𝗶𝗼𝗻𝗮𝗹 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀𝗲𝘀: 🎯 Hotels, cybersecurity firms, IT hardware companies, and consultancies often follow data center development, creating robust business ecosystems. 𝗘𝗻𝗵𝗮𝗻𝗰𝗶𝗻𝗴 𝗜𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 𝗮𝗻𝗱 𝗥𝗲𝘃𝗲𝗻𝘂𝗲: 🎯 Data centers fund significant upgrades to local roads, broadband, and power infrastructure. 🎯 They offer a high fiscal benefit, paying about $26 in local taxes for every $1 of public service used. 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝘁𝘆 𝗣𝗮𝗿𝘁𝗻𝗲𝗿𝘀𝗵𝗶𝗽𝘀: 🎯 Providers frequently support local nonprofits, educational initiatives, and community projects, contributing directly to residents' well-being. While challenges exist, communities experiencing "𝘁𝗵𝗲 𝗱𝗮𝘁𝗮 𝗰𝗲𝗻𝘁𝗲𝗿 𝗲𝗳𝗳𝗲𝗰𝘁" often see significant economic revitalization, improved infrastructure, and increased quality of life, even leading some locals to jokingly anticipate perks like getting their own Trader Joe's. 𝗖𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗔𝗿𝘁𝗶𝗰𝗹𝗲: https://lnkd.in/ecHpKgZv (𝗦𝗼𝘂𝗿𝗰𝗲: Andrew Schaap, CEO of Aligned Data Centers via Fast Company)

What if data centers didn’t just consume power but gave it back to the community... In #WestLondon, that future is being built. The #UK just approved its first large-scale data center waste heat network, led by Hemiko and backed by OPDC. The idea is simple but revolutionary: 1. Capture the excess heat from data centers 2. Pipe it into a district heating system 3. Use it to warm 9,000+ homes, schools & businesses Here’s what makes it a game-changer: • 95GWh of recovered heat • £600M in investment • 90% drop in fossil fuel heating • Powered by VIRTUS Data Centres, Ark Data Centres, Vantage Data Centers, Microsoft data centers This isn’t just a sustainability win. It’s a data center transformation story. From power-hungry black boxes. To circular infrastructure powering smart cities. Cities like #Stockholm and #Helsinki already do this. Now #London joins them, with data centers at the core. And here’s the big idea: Data centers are no longer just digital infrastructure. They’re urban infrastructure. Waste is the new resource. And smart reuse is how we build the future. #datacenters #sustainability #infrastructure

🚨NEW INVESTIGATION🚨   Naureen S. Malik Andre Tartar and I analyzed exclusive data from ~800K sensors and found a strong link between proximity to #AI #datacenters and higher levels of distorted power, which can damage home appliances and aging grids. 🎁 Read it for free here: https://lnkd.in/e4tzHbKA AI data centers are multiplying across the #US and sucking up huge amounts of power. Our analysis shows they may also be distorting the normal flow of electricity for millions of Americans. Our exclusive analysis shows that more than three fourths of highly-distorted power readings across the country are within 50 miles of significant data center activity. More than half of the tracked households showing the worst distortions of power quality are located within 20 miles of significant data center activity. US census figures show that about 3.7 million Americans live in the most-impacted areas. While the effect appears to be more severe where population density is higher, Bloomberg’s analysis shows that even in rural areas, sensors that are closer to significant data center activity are more likely to have distorted power. We visualized granular data for sensors within two Virginia communities: one located far from data center activity and recording "normal" power distortions, the other located within 7 miles from significant data center activity and recording alarming double digit readings. The latter community, in Prince William county has only experienced a glimpse of what future data center activity nearby could feel like: more than 600MW of data center capacity are rapidly being built across the highway. Our analysis sheds light on the link between significant data center activity and households with distorted power. But measuring this relationship is complex, and much more work needs to be done to address this issue. It was an absolute blast to work on this data heavy story with data whizardry Andre Tartar, incredible reporting from Naureen S. Malik and top notch edits from Chloe Whiteaker Millie Munshi Alex Newman and Brian Eckhouse 🫡 🎁 Read our investigation here: https://lnkd.in/e4tzHbKA

This Chicago Tribune story about Minooka, Illinois vividly illustrates the key dimensions of #datacenter debates underway in communities around the country, what's at stake for our #GreatLakes region, and why we need to get it right. The takeaways? - First, all parts of the Great Lakes region are not equally water rich, or water abundant. This is a common misconception. Minooka is a community in Illinois - a Great Lakes state - that is *not* water abundant. They're switching their source from Lake Michigan because their traditional source (groundwater) is running out. This was a planned switch, and an allowable one under the governance frameworks that protect the Great Lakes, but it's a dynamic worth understanding in detail when we speak about our region as water rich. - The proposed data center, as designed, will use a full 1/3 of the new water supply for cooling. Lake Michigan drinking water that takes significant energy to treat and move. As Hugh O'Hara of the Will County Governmental League asks: "Is it really the highest and best use of a finite resource to use drinkable water to cool a computer?" These are collective resource management questions we all need to grapple with. - Reducing the water consumption means increasing the energy consumption - a perfect illustration of the often overlooked energy-water connection and why surface debates about energy resources are so often about water, too. - An essential point buried deep in the piece: Equinix aims to eventually operate the entire facility with greywater (recycled wastewater) sourced from its own operations and surrounding homes and businesses. This would lessen the burden on the drinking water allocation, and allow more industrial and residential needs to be met. Why not do it today? Not technology limitations - the tech exists, and has been deployed in other countries (Current studied the utility and viability of wastewater heating and cooling for a The Water Research Foundation (WRF) project; the paper and webinars are available on our website). But there are other limiting factors: policies and regulations to support water #reuse, utility treatment capacity, and more. There is no question that water-intensive industry like data centers will be part of our region's economic future. As the article points out, they're increasingly becoming another essential utility that supports our computing-dependent lives. Our planning and policy frameworks need to evolve to support a world where we use ever drop of our precious fresh water to its highest and best use. This is a grand challenge for our region, and for the U.S. I'm grateful to be working with the support of our partners in #GreatLakesReNEW, our National Science Foundation (NSF)-backed Regional Innovation Engine, to support the transition to a circular, prosperous, and competitive #blueeconomy. https://lnkd.in/gxYuk6U8

I've written lately about how massive new data center projects for xAI in Memphis, TN and for Meta in northeastern Louisiana are being powered. Now I'd like to describe the impact that data center growth is having on a small power grid operator in Virginia. Rappahannock Electric Cooperative (REC) is a member-owned #utility that provides electric service to nearly 170,000 connections across 22 Virginia counties. It currently has zero data center load, but it's located just west of "Data Center Alley" in Virginia, the largest cluster of data centers in the world that already consumes 25% of Virginia's electricity. And Data Center Alley is expanding into REC's service territory. Today, REC's peak load is 1.1 GW, and its recently released forecast is that data center power demand alone will grow from zero today, to almost 1 GW by 2030, and to over 3 GW by 2040. Let that sink in for a moment. As a cooperative, REC is owned by its members—residential, agricultural, commercial, and industrial consumers—who receive electricity through its distribution network. It's operated as a not-for-profit entity. And rushing at it headlong are AI #hyperscalers looking to win the race to artificial general intelligence (#AGI). Alphabet Inc. (Google), Amazon Web Services (AWS), Meta and Microsoft are the four leading hyperscalers. The combined capital expenditures that they have announced for just one year (2024) to develop AI will exceed $200 billion. The cost of electricity comprises about 3-6% of the total expense of training a new AI large language model (#LLM). But in Microsoft's last earnings announcement, their mention that their business won't grow as fast as forecast because of an inability to power new data centers fast enough caused a 6% drop in their $3T market cap - that's a $180B impact. So their cost of electricity is trivial, but access to it is existential. Sleepy not-for-profit electric cooperative - meet AI hyperscaler. Sounds like a movie script. This is gonna get interesting... 

Fantastic work from Jen Liu in The Cloud is Dead series for Points Data & Society Research Institute, situating current environmental justice issues in Louisiana to histories of resistance to enslavement and petrochemical plants: "What will we be left with in the inevitable aftermath of the AI boom? While the Meta data center is slated to be in a part of the state that has not historically been a sugarcane producing area, the project has benefited from the legacies of a pro-business political environment committed to extractive economies. Living in the fallout of these previous booms has shown us that residents are often left to shoulder the risks once labor costs rise, the landscape has deteriorated, or when chemicals leach into the water and air. If Meta leaves its data center project before the plants are fully paid off, the cost will be borne by Entergy customers. Furthermore, the gas-powered plants also deviate from any plans the state has for decarbonization and broader climate goals" https://lnkd.in/e9Dn2j9g

🚀Moonshot Leadership Lesson: Why Societal Context is necessary to build tech solutions that impact millions of people In the world of moonshot building, it's easy to get caught up in the excitement of technological innovation. In fact, it's often where we spend a lot of our time. And it’s what we talk about the most. But truly transformative solutions require more than just cutting-edge tech – they demand a deep understanding of the social contexts they aim to serve and that they must emerge from. For example, the energy demands of AI are staggering. Data centers, the backbone of AI development and deployment, could consume up to 9% of U.S. electricity generation by 2030 (EPRI, 2024). This raises serious concerns about sustainability and who will benefit.  As moonshot leaders, we have a responsibility to consider the full lifecycle of our creations. That includes not just the energy consumed during development, but also the ongoing energy costs of operating AI systems at scale. The communities that are often least equipped to handle rising energy costs — those already facing economic hardship and limited resources — could be the most impacted by AI's growing energy footprint. Did you know that, “one in four low-income households have energy bills that exceed 10% of income, the level at which an energy burden is considered severe. Households facing energy burdens often face difficult choices between paying energy bills and buying other essentials, like food and medicine” (2024, ACEEE). Leading a moonshot requires a holistic perspective. We need to be as invested in understanding the social context of our work as we are in the technology itself. That means asking ourselves: How will our tech solution impact different communities, differently? Will our solutions cause harm, increase existing inequities?  How can we ensure the benefits of our innovations are experienced by those who least experience it? Part of our leadership responsibility is to champion energy-efficient AI solutions. This could involve investing in renewable energy sources for data centers, developing more energy-efficient algorithms, and promoting responsible AI practices that minimize energy consumption. Moonshots don't just solve problems through building audacious tech; they reshape the present and future with people and societal realities. By considering social context from the outset, we build smarter, more profitable, wiser and humane tech that expands rather than shrinks possibility for everyone.