How Evergreen Solar Went Belly Up

So here is what everyone thinks happened at Evergreen Solar. They asked the state for a lot of money in return for building a factory in Massachusetts, promising to bring 800 jobs to the state. They took the money, sent the jobs to China, and then declared bankruptcy, absconding with our hard earned tax dollars.

Well, that’s not quite what happened.

They were offered the $58.6 million in grants, loans, land and tax incentives by the state – they did not ask for it.

They did build a factory that employed over 800 people.

Not one job was sent to China.

They declared bankruptcy after it became clear they could not compete with heavily state-subsidized Chinese panel manufacturers who flooded the market panels they were selling, in some cases, below cost.

Some of you will see all this as a distinction without a difference, but I tell the tale here from my perspective because I was right in the middle of it when all this went down – I worked for Evergreen from April 2008 as the new plant was ramping up to March 2011 as the plant was shutting down.

The grants, loans and tax incentives were part of a deal to have the new factory built in Devens instead of elsewhere in Massachusetts or out of state because the Patrick administration wanted an anchor factory in the new Devens Industrial Park.  I believe that Westborough was one of the towns being considered by Evergreen before that deal was made, but I could be wrong about that.  Fact of the matter is that Evergreen was sufficiently capitalized to build the factory without any help from the government.

Like any other business enterprise that ends up going down the tubes – Evergreen started out with the best of intentions. They had a solar panel that used a lot less silicon, which, in theory, required a lower cost of production.  The economic model used to justify the new factory made sense in 2007 and 2008, when the company raised funds for the factory.  Solar panels were selling for about $4 per watt (the industry standard for pricing) and the forecast was that panel prices would drop to about $2 per watt in 2014.

The factory actually met all the targets.  It ramped up to building 200 megawatts of panels per year; it got the costs of panels to $2 per watt within the first 18 months of operation.  That was no small feat. The team I was on was writing the software to track the production systems as they were coming on line and putting in all sorts of long hours to get it done.  Even though the company had been around for over a decade – it very much had the feel of a start-up operation.  A lot happened very quickly.

The software systems I maintained monitored and tracked all phases of production in the factory – from raw silicon in one end to warehouse inventory of product at the other.  This position gave me an insider’s view of what was going on in the factory.  Everyone from floor supervisors and engineers to finance and management used the information we gathered.  I also wrote applications for factory workers to enter this information and programs that took production data from various machines on the floor and put it into the main database.

Two things happened about the same time that, in retrospect – would doom the company:  the Wall Street crash, and the Chinese flooding the market with their own solar panels.  Evergreen had about $30 million of stock at Lehman Brothers, which the company lost when Lehman went belly up.  Panel prices went to $2 per watt . . . in 2009.

Chinese panel manufacturers were quoted as saying they were selling their panels below cost to gain market share. Company executives went to Washington to ask for help from Congress and the Obama administration and got nowhere. 

During 2009, everyone in the company took a 10% pay cut from the CEO on down.  There was some grumbling about the fact that a floor worker making $10 per hour was getting hurt a whole lot more than a CEO pulling in a 6-figure salary by this action.

At the factory, all the employees were doing their level best to cut production costs.  Again, the company met its targets but the prices kept dropping faster.  We kept cutting production costs, increasing quality of product and meeting all production goals. Although the factory was generating cash for the company and there were buyers for every panel produced, Evergreen kept losing money. 

Evergreen had always planned to build an additional plant in Asia even before things went sour, but the economics of keeping the Massachusetts plant opened no longer made sense, although the plant here was highly automated. The Chinese government offered financial terms to Evergreen that the company just could not get in the US. What Evergreen planned to do was to build a plant in China that would make the silicon wafers then sell them to a contract manufacturer who would make them into panels – by hand. Here in Massachusetts, the plan was to continue to make wafers and turn them into solar cells to be shipped to China for assembly into panels. I even wrote the software for the tracking system to be used to inventory the cells for shipping – but it was never used.

In retrospect, Evergreen should never have announced that it was sending jobs to China. Even though they built the silicon wafer factory there – not one job was sent to China. The entire Devens operation stayed open until the end of April 2011.

Here is some irony for you. The U.S. Government, in response to complaints from the aluminum industry, slapped tariffs on aluminum products being shipped in from China. Since about 60% of the frame rails for its panels were manufactured in China – Evergreen was going to get penalized for using Chinese parts in any panel sold domestically. We spent the end of 2010 frantically programming systems to track which panels were using domestic versus foreign frames so we could know which panels had to be sold only for export.  Talk about standing in a hole and being thrown a shovel.

Still, Evergreen got the manufacturing cost of panels down to about $1.70 a watt by the beginning of 2011. It was a remarkable feat when you consider that the cost per watt was almost $3 per watt when the factory opened.  It still was not enough to make a profit and the debt just kept piling up.

When the factory closure was announced, management told us that employees would not qualify for any extra benefits due workers who lost their jobs because of foreign competition.  However, in March of 2011, I got a package from the MA Department of Labor saying that I could get extra retraining benefits because I was losing my job due to . . . foreign competition!  All I could think of when I read that was the line from the Rocky Horror Picture Show where Tim Curry says “No s--t, Sherlock”.

I was lucky though – as the economy started to recover, demand for programmers was increasing.  I started looking for a new job on January 11th, 2011 and had three offers within a month.

The day after the factory closing announcement was made in January 2011; I posted a comment on an article about the closure that was up on the WBUR web site.  In the comment I talked about how hard everyone had worked at Evergreen.  Two hours later, Curt Nickisch, the WBUR business and tech reporter, called me up to interview me and we talked for almost an hour.  Had it not been for the blizzard on January 12th, I would have been interviewed live on Morning Edition, but the blizzard was a bigger story.  I did get quoted briefly on air and on Market Place Money (my 15 seconds of fame).

After that, I got calls from the Boston Globe, the New York Times and Newsweek.  I somehow found myself being the de facto example of the laid-off Evergreen employee, despite the fact that I had it a whole lot easier than the average factory floor worker – and I never actually got laid off.

Mr. Nickisch called me again when the bankruptcy was announced later in 2011. He asked me about the way the company was managed and the quote he used on air was: 

“Could they have been better managed? From my perspective and from the perspective of a lot of people I knew who worked there; yes, they could’ve been better managed. However… I don’t think it would’ve changed the outcome.”

He also said something else in his report:

“One thing this bankruptcy filing should erase once and for all is this lingering notion that somehow Evergreen’s managers were out to fleece state taxpayers. The fact is they built a factory that got crushed by the law of supply and demand and that really pushed them into bankruptcy.”

Evergreen tried to grow quickly – too quickly, in hindsight. If they had built a smaller factory, they may have been able to ride things out.  However, who knew that the economy was going to crash like it did?  Certainly not the geniuses on Wall Street.  Who knew that the Chinese would flood the market with solar panels like they did, selling them below cost to increase market share?

Something else to think about – the Chinese Government also lost their investment in Evergreen’s Wuhan China operation, so the Chinese did not make out well either.

I know people are very unhappy because the state took a bet on Evergreen and lost.  That may be true, but other administrations (including GOP administrations) have taken bets on companies who promised to bring jobs to Massachusetts only to see those jobs get shipped out of state in spite of generous tax breaks (Fidelity anyone?).  Does that make these policies wrong?

In my opinion – yes, these policies are wrong – especially at the state and local level.  The new jobs can get people to vote for the politicians who institute the policies – but I think they don’t often get offset by sufficient tax revenue to make up for the subsidies. 

I have no great admiration for former State House Speaker Thomas Finneran –but he took a principled stand to not let the Patriots blackmail the state into subsidizing the construction of Gillette Stadium when they implied they’d move the team to Connecticut otherwise.  Finneran called their bluff and the Patriots stayed.

Finally, we as a society take bets on industries all the time or have subsidized them for so long that no one even questions it any more.  We do it for all sorts of industries. We do it for big energy (oil, coal and nuclear), we do it for agriculture, we do it for mining and we do it for defense contractors (whose waste makes the cost of the solar failures look like rounding errors).  Sometimes the bet pays off.  It did in the 1800s for the railroads; it did for the interstate highway system in the 1950s.  At this time, it is working for General Motors and Chrysler.

You can decide for yourself whether government involvement in industry makes sense at any level.

Anyway, that’s my tale.