Either labor productivity is real low in residential solar, or …
The number of solar industry jobs, now said to be 209 thousand, is widely reported and cited. President Obama included, as during his speech last April, announcing a program to train retiring military and veterans to work in solar.
The source for the number of solar jobs is an annual survey conducted by The Solar Foundation. The findings of the latest survey were published a couple of weeks ago in the "National Solar Jobs Census 2015."
The bottom line, there are 209 thousand solar jobs. Of these, 132 thousand are in residential solar.
Notably, a little more than ninety percent of these people work a hundred percent of their time on solar. Not what I would have guessed. But the survey asserts that less than one in ten do other work between installations or sales calls.
The survey is exhaustive and exhausting:
"Data for Census 2015 is derived from a statistically valid sampling and survey that went to nearly 400,000 establishments throughout the nation, in sectors ranging from manufacturing, to construction and engineering, to sales.
The results from the Census are based on rigorous survey efforts over the course of October and November 2015 that include 287,962 telephone calls and over 44,220 emails to known and potential energy establishments across the United States, resulting in data from 19,000 firms and a total of 2,350 full completions from U.S. solar establishments."
This methodology depends of course on the lack of bias among those being surveyed. However, the survey results are clearly used to boost the solar industry and encourage attractive public policies. One wonders if some bias is inherent.
So we crunched the numbers. Think of this as a sanity check, a test of whether 132 thousand residential solar jobs seems reasonable.
The survey says that 63 percent of the jobs are in residential solar, and 22 percent are in utility-scale solar (the remaining in commercial solar). Although, according to the Energy Department, 17 percent of solar megawatt-hour generation is from residential solar, and 65 percent is from utility-scale solar. In other words, each job in utility-scale produces many times more clean energy than a job in residential.
We talked last week with the survey's media contact, asking what was the breakdown among the 120 thousand solar installers, specifically how many were in residential. Having not heard back as of yet, we'll assume 63 percent of them are in residential.
This comes to about 76 thousand installers of residential solar. So of the 132 thousand residential solar jobs in total, 76 thousand are installers. The remainder, 56 thousand, are in sales of residential solar or other fields related to the residential sector.
This amount, 76 thousand, seems high for installers of residential solar. SolarCity, the market leader that installs 34 percent of all residential solar, according to GTM Research, has a little over 13 thousand employees. How can SolarCity have such a large market share with such a small workforce share?
Some number of SolarCity employees are installers, but many are in sales and other fields. It's reported that SolarCity installations are generally done by their own employees, rather than outsourced to other companies.
The implication is that the remaining 66 percent of residential solar, the market share of SolarCity competitors, is worked by nearly 67 thousand installers not employed by SolarCity. We've assumed that of all SolarCity employees, perhaps nine thousand are installers and four thousand are in other fields.
Let's assume that residential solar was installed at 300 thousand homes last year. Nobody seems to know this number for sure but if 1.5 thousand megawatts were installed and the average installation was 5 kilowatts per home, that translates into 300 thousand homes.
Looking at the residential market not addressed by SolarCity, some 200 thousand residential solar installations were worked by around 67 thousand installers. If this was the case, then there were three installations per installer in 2015.
Of course, installations involve more than one installer. Suppose generously that each residential solar installation involved four installers. Therefore, the average installer worked on twelve installations per year, with three other installers by his or her side. That's one installation per month.
One installation per month? Residential solar installations generally take a few days. What do primarily full-time installers do during rest of their time? Seems like good work if you can get it.
All this exercise shows is that the labor productivity of residential solar installations is low. Or, we're missing something.
Solar jobs are said to be increasing rapidly, at the rate of 20 percent per annum. That's inconsistent with the apparent slackness in the supply versus demand for residential installers.
We welcome any comment by The Solar Foundation or others with insight into the number of solar jobs. As space permits, we'll publish excerpts of these responses.
Steve Mitnick, Editor-in-Chief, Public Utilities Fortnightly
E-mail me: firstname.lastname@example.org