Search for how many green jobs there are and you’ll find numbers ranging from a few million to tens of millions, depending on the source. It isn’t dishonesty: nobody uses the same definition. A technician installing panels is a green job. What about whoever makes the steel for the frame? Or the accountant at the firm selling them?
In 2018 Francesco Vona, Giovanni Marin, Davide Consoli and David Popp tried to solve the problem at the root, in the Journal of the Association of Environmental and Resource Economists. Instead of classifying sectors they classified tasks: they identified, in the US occupational database, the work activities genuinely tied to environmental goals, and measured how much of each occupation consists of them.
On that measure, the share of US employment classifiable as green sits at around 2-3%. Nowhere near the press-release numbers.
Why the counts never agree
The gap between estimates isn’t random, it has a precise structure.
High estimates count people employed in sectors deemed green, including all administrative and support roles, often adding indirect employment estimated with multipliers. Low estimates count people performing environmental tasks, regardless of which sector they sit in.
They’re two different questions, and for anyone deciding what to study the second is the only useful one: it tells you what you’ll actually do all day.
There’s a second, subtler reason. An occupation can become greener without changing its name: an electrician working on photovoltaic systems is still an electrician. Alex Bowen and colleagues showed that the green transformation of the labour market consists largely of exactly this, modified content of existing occupations rather than the birth of new ones. And that moves into green jobs mostly come from very similar non-green occupations, not from distant sectors.
What green jobs actually demand
Davide Consoli and colleagues compared, in Research Policy, the skill profile of green and non-green jobs. The result is blunt and cuts against the image of green work as an easy entry point: relative to comparable occupations, green jobs require on average more formal education, more years of experience and more on-the-job training, and are more intensive in non-routine cognitive skills.
On pay, Vona, Marin and Consoli estimated across US local labour markets a premium on the order of 4% for green occupations, holding observable skills constant. That’s a real but modest premium: it doesn’t on its own justify uprooting your life, while it very much justifies steering an existing career in that direction.
On location, a study based on job postings showed that solar and wind positions are a small fraction of total postings, are growing, and are heavily concentrated geographically. Most aren’t research roles: they’re installation, construction, maintenance and sales.
| Common belief | What the data say |
|---|---|
| Green jobs are millions of new positions | Around 2-3% of employment on task-based measures |
| You need an environmental degree | Experience and technical training matter more than specific degrees |
| They’re lab and research jobs | Mostly installation, maintenance, site work, sales |
| They’re everywhere | Heavily concentrated in a few geographic areas |
| They pay much more | Estimated premium around 4% holding skills constant |
The part nobody tells you: who pays for the transition
Here’s the hardest finding in the whole literature, from Reed Walker in the Quarterly Journal of Economics. Using individual administrative data, Walker followed workers in industries affected by US air quality regulation.
The result: those working in the affected industries suffered substantial earnings losses in the following period, on the order of a large fraction of a year’s salary in present value, driven mainly by spells of non-employment and by lower wages in the new sector.
On the other side, David Popp and colleagues studied the large US green stimulus of 2009: there were employment effects, but they appeared with a lag of years, and were concentrated in areas that already had a stock of green skills and in manual, non-degree roles.
The transition creates jobs and destroys others. The problem is that the two groups of people rarely coincide, and the bill is paid by the first group before the second one exists.
That’s the same mechanism described in jobs of the future, and the individual cost of switching sector is quantified in reskilling.
How to actually move into a green job
- Don’t change sector: green the one you’re in. The route documented as most frequent is a move from one occupation to a very similar one with environmental content, not a leap between worlds.
- Look at where the installations are, not where the news is. Geographic concentration is strong: an hour spent researching active projects in your area beats ten national articles.
- Aim at on-site technical trades. Installation, maintenance, efficiency retrofits: non-offshorable tasks that won’t be automated soon.
- Learn the regulatory side. Reporting, certifications, disclosure obligations: scarce, domain-specific skill that lasts as long as the rule does.
- Be sceptical of generic sustainability certificates. The signal that counts in the data is verifiable technical experience, not the certificate.
- Check whether your sector is exposed on the wrong side. If you work in a supply chain the transition hits, the time to move is while you still have a job, not afterwards.
- Reason about tasks, not labels. Same logic I use for choosing skills in the skills of the next five years.
My take: inflating the numbers bred distrust
I think green-jobs communication has done avoidable damage by inflating the numbers. When the promise is “millions of jobs” and the reality is “a small but real share, concentrated, with high technical requirements”, the result isn’t enthusiasm: it’s distrust of the entire topic, which actually rests on solid ground.
I also think Walker’s result deserves a place in every public discussion of the transition, and almost never gets one. A policy can be right in aggregate and devastating for specific people, and those two things don’t cancel out.
One idea the data can’t confirm yet: over the coming years the bottleneck in the transition won’t be technology or money, but the number of people able to physically install and maintain things. It’s the least-discussed trade and the one where demand most visibly exceeds supply.
Frequently asked questions
Do I need an environmental degree? For most roles, no. The data show technical experience and on-the-job training matter more than a specific qualification.
Do green jobs pay more? A little, on the available estimates, holding skills constant. The premium is measurable but not huge.
Which roles have the most demand? On-site technical ones: installation, maintenance, energy efficiency, plus regulatory compliance and reporting roles.
Is it a bubble? The part driven by regulatory obligations and investment already committed is the least volatile. The part driven by communications and corporate image is far more so.
Where do I start if I come from another sector? From the contact point between your sector and the transition, not from the centre of the topic: that’s the trajectory with the lowest switching cost.
Sources
- Vona, F., Marin, G., Consoli, D., & Popp, D. (2018). Environmental Regulation and Green Skills: An Empirical Exploration. Journal of the Association of Environmental and Resource Economists, 5(4), 713-753. The task-based measure of green skills.
- Vona, F., Marin, G., & Consoli, D. (2019). Measures, Drivers and Effects of Green Employment: Evidence from US Local Labor Markets, 2006-2014. Journal of Economic Geography, 19(5), 1021-1048. The green employment share and the wage premium.
- Consoli, D., Marin, G., Marzucchi, A., & Vona, F. (2016). Do Green Jobs Differ from Non-Green Jobs in Terms of Skills and Human Capital? Research Policy, 45(5), 1046-1060. Education, experience and training required.
- Bowen, A., Kuralbayeva, K., & Tipoe, E. L. (2018). Characterising Green Employment: The Impacts of “Greening” on Workforce Composition. Energy Economics, 72, 263-275. The transformation of existing occupations.
- Walker, W. R. (2013). The Transitional Costs of Sectoral Reallocation: Evidence from the Clean Air Act and the Workforce. Quarterly Journal of Economics, 128(4), 1787-1835. Earnings losses for workers in regulated industries.
- Popp, D., Vona, F., Marin, G., & Chen, Z. (2021). The Employment Impact of a Green Fiscal Push: Evidence from the American Recovery and Reinvestment Act. NBER Working Paper No. 27321. The multi-year lag in employment effects.
- Curtis, E. M., & Marinescu, I. (2022). Green Energy Jobs in the US: What Are They, and Where Are They? NBER Working Paper No. 30332. The posting-based measure and geographic concentration.
- Autor, D. H., Levy, F., & Murnane, R. J. (2003). The Skill Content of Recent Technological Change: An Empirical Exploration. Quarterly Journal of Economics, 118(4), 1279-1333.
- Jacobson, L. S., LaLonde, R. J., & Sullivan, D. G. (1993). Earnings Losses of Displaced Workers. American Economic Review, 83(4), 685-709. The classic reference on the individual costs of reallocation.
- International Renewable Energy Agency and International Labour Organization. Renewable Energy and Jobs. Annual series. Source of the high, sector- and multiplier-based estimates, useful for understanding where the discrepancy comes from.
This article was written following the process described in how we write. If you find a figure reported wrongly, write to me: I’ll correct it and note the correction at the bottom of the article.



