Proponents of clean energy have long argued that investment in solar, wind, and other renewable sources creates domestic jobs. In the past few years, however, critics of renewable energy have responded — with considerable success — by arguing that the net effect is actually negative.
The concept of “netting” the effect of investments (including government subsidies) on the job market is legitimate when used correctly; when not used correctly, it is disingenuous.
Take the case of the F-22 fighter plane. You may have heard the current debate over the decision to discontinue production of this aircraft. One argument against the proposal held that jobs in the aviation production sector would be lost. The other camp countered with the net-effect argument: we are talking about the allocation of a portion of the military budget, not additional funding for new F-22s, they pointed out, so the question is whether the same amount of money could be better used on other things — including on other aircraft. To the extent that other planes can be built by the same companies that would manufacture the F-22, the net job effect may be negligible. Aircraft will be built anyway. We are comparing the production of one plane to the production of others.
With renewables, the question of net effect is also related to other, similar investments. Since renewables generally produce electricity, not investing in them would mean investing in electricity from other sources. What, then, are the effects on the labor market when we invest in renewables rather than in fossil fuels or nuclear?
I would love to give you the answer to that question, but to my mind, few sound studies have ever been conducted on the issue (I note one exception below) — which has not stopped a growing number of critics of renewables from using the net-effect argument.
The study that’s received the most attention is entitled “Study of the effects on employment of public aid to renewable energy sources” (PDF), made available March of 2009 by Professor Gabriel Calzada Álvarez and his team at Madrid’s Rey Juan Carlos University as a “draft: bibliography pending.” Interestingly, as of this writing at the end of August, we are still waiting on the bibliography.
Although the paper is merely a draft and did not undergo any peer review, it has received media attention at the highest level. “Saving the planet and creating jobs may be incompatible,” wrote the British Economist in a review of the study. In the U.S., Fox News surmised: “Americans shouldn’t be depending on green jobs to help the U.S. economy.” And the New York Times wrote, “Alvarez posits that the programs creating those jobs resulted in the destruction of nearly 110,000 jobs elsewhere in the economy.” Word has it that the paper is being circulated in the US Congress; it was also cited favorably in a Michigan white paper opposing the implementation of European feed-in tariffs (FITs) in the US.
The same net-effect argument has also been used by a group of economists associated with the Cato Institute, led by Prof. Roger Meiners of the University of Texas at Arlington. In the last year, the group has published a 97-page report “Green Jobs Myths” and the shorter (21 pages) “7 Myths about Green Jobs,” both of which aim to prove that almost nothing said positively about renewables and job creation is true.
The Spanish and the US reports share a common flaw: they do not say what would otherwise be done with the money invested in renewables. But remember the F-22 decision above — you cannot talk about the net effect unless you say what the alternatives are.
When Álvarez does mention an option, he unwittingly uses an example most Americans consider unattractive: the “unseen cost” of renewables is “all the hamburgers not cooked … as a result of the state directing resources to windmills or solar panels” — a nonsensical example (are we supposed to believe that Spain does not have enough hamburgers?) that he may want to reconsider, for “flipping burgers” is synonymous in the US with poorly paid, unskilled labor.
The formula used in these reports is astonishingly simplistic: take the amount invested in a sector and divide it by the number of jobs created; you get the cost of each job. Now take overall investments in the general economy and divide that figure by the number of jobs created overall, and you get the cost of each job in the general economy.
Álvarez found that 571,138 euros is invested in each green job in Spain, compared to only 259,000 euros per job in the general Spanish economy. He therefore concludes that 2.2 jobs are not created for each green job. I have no idea if these figures are accurate, because he does not provide his sources or define his terms, but no matter — we will take him at his word. The argument collapses all the same.
If we compare Meiners’ figures with those from Spain, we have some interesting results. Meiners and his colleagues put the figure at “$107,000 per new job” in the renewables sector. At this point, a disingenuous critic of Meiners and Álvarez might point out the discrepancy in their figures and calculate that the cost of a green job calculated by Meiners offsets only 0.3 jobs at the general costs calculated by Álvarez. In other words, you create three more jobs in the renewables sector than you would in the overall economy. But throwing together figures calculated in different ways is not honest scholarship. Nonetheless, Álvarez and Meiners cobble together their statistics in this very way.
To their credit, Meiners and his colleagues highlight the lack of a clear definition of the term “green jobs.” Unfortunately, they do not define what they mean by it and use that definition to collect and compare a fresh set of statistics. Rather, they cite a wide range of other papers and studies — with no attempt to adjust for the discrepancies behind all of the figures.
Since the work of Álvarez and his colleagues has not been peer-reviewed, I did them the favor on behalf of Germany’s photovoltaik magazine, which published the peer review. Of the four reviewers asked to participate, three said the paper was not sound; the dissenting voice was Meiners himself, who said he supported publication of the paper because, “I oppose censorship.” Unsurprisingly, Meiners seemed pleased with the paper’s approach — which, indeed, is his own: “If we push billions of euros in one direction rather than other, something else must be given up. So the explanation made sense.” Meiners agreed that the calculations in the paper were not explained in full, but that did not bother him: “as usual, we presume people do not lie as their reputations are on the line.”
Another reviewer was Jochen Diekmann, deputy head of the Department of Energy, Transport, and the Environment at the German Institute of Economic Research (DIW), a “leading research institute involved in basic research and policy advice.” Dr. Diekmann’s response sums up those of the other two reviewers who opposed publication: “The study is not based on the researchers’ own calculations about labor market effects,” but rather on previously published figures, such as from the EU’s MITRE study. But this combination concerned Diekmann because, “In all likelihood, combining figures from different scenarios and time frames leads to inconsistencies, which is generally also the case for the combination of MITRE results from 2003 and the more recent data the study uses elsewhere.”
The very wording of these two responses — one from an American and one from a German — are revealing. The American produces complete non sequiturs (What does censorship have to do with a peer review? And what planet do you have to live on not to realize that people whose reputations are on the line lie all the time?), while the German produces objective, dispassionate criticism that slices right to the heart of the matter.
No wonder the best study, to my knowledge, on the net job effects of investments in renewables is from Germany. In September 2006, the German Ministry of the Environment, Nature Conservation, and Nuclear Safety published a study entitled “Impact of the Expansion of Renewable Energy on the German Labour Market” (PDF), which concluded that “the net impact … is also a clear and sustainable positive employment stimulus.”
Other studies in the US — as long as we are throwing around disparate figures — support the idea that a lot of well paying jobs will be created by renewable investments. Robert Pollin, co-director of the Political Economy Research Institute at the University of Massachusetts, has estimated that most “green” (again, definitions vary) workers — from engineers to construction workers to marketing staff — will earn more than 20 US dollars per hour on the average. At that rate, you might actually need 2.2 jobs flipping hamburgers to earn as much.
Perhaps such studies could be considered biased; in each case, the conclusions were amenable to the authors’ politics. But have the statistics been calculated or collected to produce a desirable outcome? The author of the Michigan White Paper mentioned above has accused me of “pre-existing bias” — a strange formulation, as though an ex post facto bias would be preferable.
A person is not biased simply because they disagree with you. They may have reached their conclusions based on the facts. In contrast, you are biased if you get the facts wrong in order to strengthen your argument — which the author of the Michigan White Paper does when he writes that, “Germany and Spain have had huge percentage increases in solar energy production, but solar accounts for less than 1% in both countries after 15 years of FIT [feed-in tariff] subsidies.” Actually, Germany started offering its FITs less than nine years ago — but no matter, 15 years it sounds better if you want to argue that the subsidies are of no avail. What do the facts matter?
While Álvarez and Meiners are both free marketeers who believe that the market knows best, the Spaniard writes that Spain’s support for renewables will leave the country “saddled with and further artificially perpetuating obsolete fixed assets, far less productive than cutting-edge technologies.” So governments do not know which technologies will prevail, but economists do? Meiners seems to think so: “Some technologies preferred by the green jobs studies are not capable of efficiently reaching the scale necessary to meet today’s demands.” Which ones? How does he know? Why doesn’t he tell us? Do Álvarez and Meiners base their assumptions about how the market gets everything right on the financial sector’s performance over the past four quarters?
In “7 Myths About Green Jobs,” Meiners and his colleagues correctly write, “Wind plays an increasing role in electricity generation, but electricity is only a fraction of energy production in the U.S., which is why wind is such a tiny share of energy” (italics in original). But only one page later, we read that nuclear “produces about 20% of U.S. electric power” — with no further reminder that nuclear also only produces electricity.
Álvarez’ study, allegedly about economics, contains much that is outside of economics — and false. He writes of wind, “The rate of development of this technology has remained comparatively quite calm,” though he does not explain what he means. Is he not aware that standard turbines have grown from 55 kilowatts to more than 5,000 kilowatts in two decades? He writes elsewhere, “solar failed even to reach 1% of Spain’s total electricity production in 2008.” But solar grew by around 100 percent in Spain both in 2007 and 2008. So next year, would he have complained that it failed to reach 2 percent had Spanish policy not been changed?
Finally, while the Spanish study quotes the CEO of Acerinox complaining that retail electricity prices had increased by 10.6 percent in just two years, that quote comes from April 2002. At the time, Spain only had a fraction of its current installed wind capacity — and practically no photovoltaics to speak of. Back then, renewables could not have been the main culprit. And Acerinox left Spain for a “free economic zone” in Indonesia, where it can escape much more than high energy costs.
Most tellingly, the net-effect argument proves to be a dismal failure when it is actually applied. For instance, one semiconductor manufacturer has set up several production plants in Europe and has just announced that one will be built in the US. For one of its European plants, it received 545 million euros in subsidies; eventually, some 1,000 jobs were created there, putting the cost of each job at 545,000 euros according to the simplistic calculation commonly used in the Spanish paper. But we now see that, by the same simplistic formula, investments in the semiconductor sector also offset some 2.2 jobs (if the figures for Spain roughly hold true for Germany). And the figures are getting worse: the company’s new U.S. plant will employ some 1,400 people and receive 1.2 billion dollars in subsidies — some 857,000 dollars per job.
Here, it becomes clear how much the Spanish study leaves out. Does the value of the semiconductors from those plants make these subsidies seem negligible? What does the state stand to get back in corporate and other taxes? If local jobs are created and foreign jobs offset, the overall effect is positive anyway. Renewables offset foreign resources in most cases. Spain gets roughly half of its energy from coal, but has imported more than it produces domestically for years. The situation is similar in Germany, and energy independence is a buzz word in the US. But the Spanish study does not take domestics jobs offsetting foreign jobs into account.
An estimate of the net effect of investments in renewables might be enlightening, but it would require, as Meiners points out, an agreed definition of what constitutes a green job. It would also, however, have to include take into account how the investments are spent otherwise and how we are going to get the amount of electricity that would have come from renewables. Anything else is twisting the facts.
But my favorite calculation is this one: the last annual budget of the University of Rey Juan Carlos, where Álvarez teaches, was 22,889,932.93 euros. With a staff of around 2,000 people, that comes to some 416,180 euros per job, according to his own method. In other words, Prof. Álvarez’s calculation would seem to suggest that shutting down his university would create 1.5 jobs (416,000 euros divided by 259,000 euros per job in the general Spanish economy) for each job lost. I wonder if Mr. Álvarez would prefer, to take the example he gave, flipping burgers 60 hours a week to his cushy job that allows him to upload draft papers to the web.
These days, it seems that misleading information is a legitimate option in public debate in the US (it is not in Germany). Until I see a sound study to the contrary, I will view the net-effect argument right up there with “death panels” and Obama’s Kenyan birth: a nefarious attempt at winning by any means possible.