A study of 1,584 U.S. data centers found job growth in supporting industries, but little boost for local restaurants and shops or in jobs at the data centers themselves.
Two professors who study real estate and urban economics, including how large place-based investments shape local economies, set out to answer a question at the center of these fights: Do data centers actually create jobs where they are built?
"During the first two years after a county’s first data center opened, we found little change in employment outside construction. After that, jobs begin to grow. Three to five years after a center opened, we found that these counties had roughly 1,000 to 1,300 additional nonconstruction jobs than comparable counties without data centers."
... "Our findings do show that data centers can create jobs beyond the small number of people working inside their walls. But that does not mean every one of them is a good deal for their local community. And whether those additional jobs justify the costs remains an open question."
Opponents of large data centers often use the projected local job production (not counting initial construction) which is relatively low compared to industries such as manufacturing as leverage in their efforts to advocate against these facilities i.e. the job benefits would not be worth the supposed downsides in local utility usage. Yet, this issue of employment becomes somewhat arbitrary. What level of employment potential justifies new local industrial buildouts? This article hints at the potential for data centers to create several thousand additional jobs (without mentioning the potential downstream economic benefits from those additional jobs) after a few years.
However, focusing on a few thousand local jobs may be missing the forest for the trees. According to the US Bureau of Labor Statistics, the biggest job creator after healthcare, education, and retail is manufacturing at number 4 (non-manufacturing high-tech jobs like in data centers is 8 on the list). Given the propensity for rapid advancements in high tech including more efficient/self-cooling chips, improved "AI" efficiencies, etc. it is plausible that the infrastructure needed to support data centers could shrink dramatically over the next decade leading to a broad downsizing of these centers. This would leave relatively new space and infrastructure to support new manufacturing endeavors. Basically, the mad-rush of the tech industry to build these centers could - in effect - be subsidizing the creation of infrastructure to support the growth of local manufacturing and this includes new power generation and this has the potential for significant local job growth.
Should there not be as much in the way of down-sizing of data center foot prints then in the very least the initial facility and utility buildout could serve as an anchor industry for the expansion of an even larger manufacturing or other high tech industrial park and that is a whole lot more jobs.
Governments should consider this economic potential when debating the pros and cons of local data centers. Instead, governments can build this potential into legislation requiring the construction of supporting power generation with local interconnection that could benefit the community from surplus power, efficient reclaimed water systems, and pad construction for future industrial and manufacturing expansion instead of such staunch opposition.
>which is relatively low compared to industries such as manufacturing
This isn't necessarily true. Even according to the National Association of Manufacturers around 75% of manufacturing firms had fewer than 20 employees and over 90% had fewer than 100, and this is entire firms not just the people on-site, whereas most statistics on datacenter job creation only focus on the ones physically in the datacenter, and not the Network Engineers, DevOps engineers, Software engineers, etc who have jobs at the same company because of the datacenter.
It depends a lot on what is manufactured, but it often doesn't take that many more more people if not the same amount to watch over an automated system of robots and conveyer belts than it does to maintain a few hundred thousand servers.
We'll never "bring manufacturing back" like many people want until we get people to understand/accept that manufacturing is not a jobs program.
Even for vehicle manufacturing. The F-150 assembly factories employ 10,000 people across two factories in America. That's less than many office buildings in New York City for America's best selling vehicle.
If they want a factory near them, they will likely get about the same economic benefit.
The only full time job in a data centre is the security guard. And by full-time I mean one FTE with the security guards all working part time and making low pay and not working enough hours to qualify for benefits. Everyone else coming and going is likely a vendor or contractor in to do some quick work or scheduled maintenance and then leaves.
>The only full time job in a data centre is the security guard.
This is not true. All the infrastructure in one is under strict SLAs so they need people on site 24/7 to fix anything that comes up.
Plus, a typical hyperscaler datacenter has about 100,000 servers at least. With any system with that many individual parts, there's always something breaking, so they have people 24/7 replacing parts. Even if 10% of ram sticks fail before EOL, with 12-36 sticks per server and 100,000 servers that means that things are regularly breaking.
Opponents of large data centers often use the projected local job production (not counting initial construction) which is relatively low compared to industries such as manufacturing as leverage in their efforts to advocate against these facilities i.e. the job benefits would not be worth the supposed downsides in local utility usage. Yet, this issue of employment becomes somewhat arbitrary. What level of employment potential justifies new local industrial buildouts? This article hints at the potential for data centers to create several thousand additional jobs (without mentioning the potential downstream economic benefits from those additional jobs) after a few years.
However, focusing on a few thousand local jobs may be missing the forest for the trees. According to the US Bureau of Labor Statistics, the biggest job creator after healthcare, education, and retail is manufacturing at number 4 (non-manufacturing high-tech jobs like in data centers is 8 on the list). Given the propensity for rapid advancements in high tech including more efficient/self-cooling chips, improved "AI" efficiencies, etc. it is plausible that the infrastructure needed to support data centers could shrink dramatically over the next decade leading to a broad downsizing of these centers. This would leave relatively new space and infrastructure to support new manufacturing endeavors. Basically, the mad-rush of the tech industry to build these centers could - in effect - be subsidizing the creation of infrastructure to support the growth of local manufacturing and this includes new power generation and this has the potential for significant local job growth.
Should there not be as much in the way of down-sizing of data center foot prints then in the very least the initial facility and utility buildout could serve as an anchor industry for the expansion of an even larger manufacturing or other high tech industrial park and that is a whole lot more jobs.
Governments should consider this economic potential when debating the pros and cons of local data centers. Instead, governments can build this potential into legislation requiring the construction of supporting power generation with local interconnection that could benefit the community from surplus power, efficient reclaimed water systems, and pad construction for future industrial and manufacturing expansion instead of such staunch opposition.
[OP] StormyTrumpy | 19 hours ago
A study of 1,584 U.S. data centers found job growth in supporting industries, but little boost for local restaurants and shops or in jobs at the data centers themselves.
Two professors who study real estate and urban economics, including how large place-based investments shape local economies, set out to answer a question at the center of these fights: Do data centers actually create jobs where they are built?
"During the first two years after a county’s first data center opened, we found little change in employment outside construction. After that, jobs begin to grow. Three to five years after a center opened, we found that these counties had roughly 1,000 to 1,300 additional nonconstruction jobs than comparable counties without data centers."
... "Our findings do show that data centers can create jobs beyond the small number of people working inside their walls. But that does not mean every one of them is a good deal for their local community. And whether those additional jobs justify the costs remains an open question."
jestina123 | 16 hours ago
>By contrast, we found little evidence of job growth in restaurants, stores
Data centers usually have less than 500 permanent employees. Incredible journalism.
DartosMD | 17 hours ago
Opponents of large data centers often use the projected local job production (not counting initial construction) which is relatively low compared to industries such as manufacturing as leverage in their efforts to advocate against these facilities i.e. the job benefits would not be worth the supposed downsides in local utility usage. Yet, this issue of employment becomes somewhat arbitrary. What level of employment potential justifies new local industrial buildouts? This article hints at the potential for data centers to create several thousand additional jobs (without mentioning the potential downstream economic benefits from those additional jobs) after a few years.
However, focusing on a few thousand local jobs may be missing the forest for the trees. According to the US Bureau of Labor Statistics, the biggest job creator after healthcare, education, and retail is manufacturing at number 4 (non-manufacturing high-tech jobs like in data centers is 8 on the list). Given the propensity for rapid advancements in high tech including more efficient/self-cooling chips, improved "AI" efficiencies, etc. it is plausible that the infrastructure needed to support data centers could shrink dramatically over the next decade leading to a broad downsizing of these centers. This would leave relatively new space and infrastructure to support new manufacturing endeavors. Basically, the mad-rush of the tech industry to build these centers could - in effect - be subsidizing the creation of infrastructure to support the growth of local manufacturing and this includes new power generation and this has the potential for significant local job growth.
Should there not be as much in the way of down-sizing of data center foot prints then in the very least the initial facility and utility buildout could serve as an anchor industry for the expansion of an even larger manufacturing or other high tech industrial park and that is a whole lot more jobs.
Governments should consider this economic potential when debating the pros and cons of local data centers. Instead, governments can build this potential into legislation requiring the construction of supporting power generation with local interconnection that could benefit the community from surplus power, efficient reclaimed water systems, and pad construction for future industrial and manufacturing expansion instead of such staunch opposition.
MajesticBread9147 | 16 hours ago
>which is relatively low compared to industries such as manufacturing
This isn't necessarily true. Even according to the National Association of Manufacturers around 75% of manufacturing firms had fewer than 20 employees and over 90% had fewer than 100, and this is entire firms not just the people on-site, whereas most statistics on datacenter job creation only focus on the ones physically in the datacenter, and not the Network Engineers, DevOps engineers, Software engineers, etc who have jobs at the same company because of the datacenter.
It depends a lot on what is manufactured, but it often doesn't take that many more more people if not the same amount to watch over an automated system of robots and conveyer belts than it does to maintain a few hundred thousand servers.
We'll never "bring manufacturing back" like many people want until we get people to understand/accept that manufacturing is not a jobs program.
Even for vehicle manufacturing. The F-150 assembly factories employ 10,000 people across two factories in America. That's less than many office buildings in New York City for America's best selling vehicle.
If they want a factory near them, they will likely get about the same economic benefit.
cvr24 | 17 hours ago
The only full time job in a data centre is the security guard. And by full-time I mean one FTE with the security guards all working part time and making low pay and not working enough hours to qualify for benefits. Everyone else coming and going is likely a vendor or contractor in to do some quick work or scheduled maintenance and then leaves.
MajesticBread9147 | 16 hours ago
>The only full time job in a data centre is the security guard.
This is not true. All the infrastructure in one is under strict SLAs so they need people on site 24/7 to fix anything that comes up.
Plus, a typical hyperscaler datacenter has about 100,000 servers at least. With any system with that many individual parts, there's always something breaking, so they have people 24/7 replacing parts. Even if 10% of ram sticks fail before EOL, with 12-36 sticks per server and 100,000 servers that means that things are regularly breaking.
Alternative-One8359 | 4 hours ago
So your study showed that data centers are extremely beneficial for counties with low populations.
Alternative-One8359 | an hour ago
I have a ranch in a county with a 5k population. I wonder what 1k additional jobs would do for that county....
Alternative-One8359 | 4 hours ago
So your study shows data centers have an extremely positive job impact on counties with low populations.
DartosMD | 17 hours ago
Opponents of large data centers often use the projected local job production (not counting initial construction) which is relatively low compared to industries such as manufacturing as leverage in their efforts to advocate against these facilities i.e. the job benefits would not be worth the supposed downsides in local utility usage. Yet, this issue of employment becomes somewhat arbitrary. What level of employment potential justifies new local industrial buildouts? This article hints at the potential for data centers to create several thousand additional jobs (without mentioning the potential downstream economic benefits from those additional jobs) after a few years.
However, focusing on a few thousand local jobs may be missing the forest for the trees. According to the US Bureau of Labor Statistics, the biggest job creator after healthcare, education, and retail is manufacturing at number 4 (non-manufacturing high-tech jobs like in data centers is 8 on the list). Given the propensity for rapid advancements in high tech including more efficient/self-cooling chips, improved "AI" efficiencies, etc. it is plausible that the infrastructure needed to support data centers could shrink dramatically over the next decade leading to a broad downsizing of these centers. This would leave relatively new space and infrastructure to support new manufacturing endeavors. Basically, the mad-rush of the tech industry to build these centers could - in effect - be subsidizing the creation of infrastructure to support the growth of local manufacturing and this includes new power generation and this has the potential for significant local job growth.
Should there not be as much in the way of down-sizing of data center foot prints then in the very least the initial facility and utility buildout could serve as an anchor industry for the expansion of an even larger manufacturing or other high tech industrial park and that is a whole lot more jobs.
Governments should consider this economic potential when debating the pros and cons of local data centers. Instead, governments can build this potential into legislation requiring the construction of supporting power generation with local interconnection that could benefit the community from surplus power, efficient reclaimed water systems, and pad construction for future industrial and manufacturing expansion instead of such staunch opposition.