Last week, the U.S. Energy Information Administration published a report claiming that only 30 percent of distributed solar is owned by third parties, with a state-by-state breakout for major states by segment. The data looks great, especially for purveyors of direct ownership models -- except it’s wrong. Irresponsibly wrong.

It’s not that this is just another chapter in EIA’s misreporting of U.S. solar’s current and projected contribution to the U.S. energy mix; it’s that this data potentially misleads solar entrepreneurs and corporates considering how to build, maintain, manage and otherwise invest in gigawatts of new and existing distributed solar.

We won’t quibble on EIA’s denominator of 12.3 gigawatts for distributed solar. For the record, we’ve tracked 14.7 gigawatts (DC) of distributed solar through Q3 2016 (via SEIA/GTM Research’s U.S. Solar Market Insight Q4 2016), which converts to ~12.8 gigawatts (AC) via a back-of-envelope 0.87 DC-to-AC ratio. However, EIA’s 30 percent third-party ownership figure representing 3.7 gigawatts doesn’t even pass an eyeball test.

Consider that SolarCity, Sunrun and Vivint Solar have together installed/deployed 4 gigawatts (DC) (~3.5 GW AC). Of course, some of these deployments have been cash or loan, not PPA or lease, but even this discounting would mean that the rest of the third-party ownership (TPO) market would be confined to just a few hundred megawatts. That’s before we add in contributions from other major deployers of third-party solar like Sungevity, Sunnova, SunPower, SunEdison, NRG, Borrego, etc., over the years.

Looking at EIA’s state-level breakouts should raise similar red flags. EIA’s findings show that only 11 percent of commercial and industrial (C&I) solar customers utilize a TPO structure. That means less than 500 megawatts and 0 (!!!) TPO C&I in California and Massachusetts, respectively. It’s clear that there’s something amiss with the data set. 

For the record, our estimates categorize roughly 7.9 gigawatts (DC) (~6.9 GW AC) of distributed solar as TPO from 2011 through Q3 2016. Even without estimating pre-2011 TPO totals, that figure indicates that over half of the distributed solar installations are served by leases or PPAs. While we’ll agree the industry is well past peak TPO for the residential sector, PPAs and leases still represent more than half of all new residential solar installations today. Furthermore, third-party ownership structures made up two-thirds of U.S. non-residential solar in 2015 according to our research on the U.S. commercial solar landscape.

As solar companies, investors and utilities look toward new business models for deployment, asset management, grid integration or even plant aggregation, the industry needs an accurate account of who owns what and how much of it they own in order to make the best decisions.

Tracking the U.S. solar market is tough. We get it. We do it every quarter by knocking on the metaphorical doors of more than 100 utilities, regulators, incentive programs and installers just for the dataset that feeds into U.S. Solar Market Insight reports. We’ve collected some form of project-level data on over 1,000,000 PV systems installed in the U.S., as well as survey-level responses that represent an additional couple of hundred thousand systems. It’s not easy to check, clean and compile all these disparate data sources, and we’ve certainly made mistakes, too. So we applaud EIA’s efforts to gather and publish more data on distributed solar. In fact, we wish they’d get better sooner so our analysts can spend their time analyzing the market, not harassing utility interconnection programs with data requests.

But the EIA needs to recognize when their reports based on incomplete data misrepresent the solar industry at such an obvious level. Good data representation should be the least we can expect from the U.S. federal government’s principal agent of energy information. We’re data nerds, too, so we’d be happy to collaborate and work with the EIA and others on building more accurate data sets. Better data means better analysis and more informed decision-making, which is good for all energy industry stakeholders.