Is the U.S. government keeping tens of millions of people poor by focusing its development assistance on renewables rather than gas and coal? It’s a critical question – particularly as the United States ramps up its Power Africa effort – that’s addressed thoughtfully by Todd Moss and Benjamin Leo in a new Center for Global Development paper that Bjorn Lomborg highlighted in a USA Today column this weekend.
Moss and Leo estimate that OPIC (Overseas Private Investment Corporation) investment restrictions, which tightly cap the amount of money that can go toward coal or gas, are costing 60 million people access to electricity. But while I’m on board with their basic point – the world’s wealthy should avoid cutting greenhouse gas emissions on the backs of poor people when rich ones are still pumping out so much pollution – I’m skeptical of their bottom line.
The Moss and Leo argument is straightforward. Every dollar committed by OPIC to a natural gas power project is accompanied by four dollars (on average) in private funds. In contrast, every OPIC dollar committed to a renewable energy project leverages only 50 cents on average. If OPIC has a ten billion dollar portfolio, dedicating that to gas would generate 50 billion dollars in investment, but directing it toward renewables would yield only 15 billion dollars. Moss and Leo combine these figures with an estimate of the per-person costs of energy access to conclude that focusing on gas could generate access for 90 million people but that investing in renewables would serve only 20-27 million. They also estimate the amount of generating capacity that each type of focus could yield: only 4,200 megawatts (MW) for renewables but 42,000 MW for one-third-less-capital-intensive natural gas.
It strikes me that there are three issues with this analysis. Correcting two of them makes the trade-off look less stark, but fixing the third makes it look even worse.
The first problem is with the leverage ratios. The historical leverage ratios do not tell us that for every additional dollar OPIC spends on gas the private sector will spend four. They actually tell us nothing about how much private investment a dollar of OPIC spending will leverage, because they don’t tell us what happens at the margin, and they don’t tell us anything about causality. To see why, imagine that private investors planned to spend a billion dollars on natural gas. Now imagine that OPIC stepped up and decided to commit ten million dollars to the same end – and that developers pocketed that money without expanding their projects. We would calculate a stunning 100:1 leverage ratio for this project, even though the actual leverage is zero. It’s entirely possible that public spending on natural gas projects appears to leverage more private capital than spending on renewables does simply because more private capital is already there for natural gas than for renewables. We have no idea, at least based on the numbers that Moss and Leo present, whether OPIC investment attracts more private capital when it’s in gas or in renewables.
The second issue has to do with how costs are defined. Moss and Leo focus on power plant capital costs. Those are, to a good approximation, the full costs of renewables. But they also are, of course, far from the full costs of natural gas. (Generation costs for natural gas are typically dominated by the cost of fuel.) Now Moss and Leo note that most of the countries targeted by Power Africa have decent natural gas resources. But there is a cost to using these domestically: foregone export revenues. (And there’s a cost to producing them, namely the labor and capital – probably imported in the latter case – that’s required.) It’s not much use to build gas-fired generating capacity unless there’s affordable fuel for it to use (just ask Indian planners). At a minimum, then, you’d need to argue that consumers will be able to pay for the continued operation of OPIC-backed gas-fired power plants. To be more complete you’d need to look at the full cost of gas-fired generation in any comparison – including foregone investment or revenues resulting from more domestic gas use.
There is, however, a third thing that weighs strongly toward Moss and Leo’s bottom line: not all megawatts are created equal. A megawatt of wind or solar doesn’t deliver nearly as much electricity over time as a megawatt of gas-fired capacity can. Assuming that fuel is available at an affordable price – a significant assumption that we’ve just looked at – the 42,000 MW of gas-fired capacity that Moss and Leo estimate are actually more than 10 times as valuable as the 4,200 MW of renewables they flag.
What’s the bottom line here? Moss and Leo are right to warn us against shortchanging the poor by being dogmatically opposed to supporting fossil fuel development. But it’s far from obvious that directing OPIC money toward natural gas projects consistently yields bigger energy access payoffs than spending it on zero-carbon electricity. Better to pursue a project-by-project assessment of costs and benefits that focuses on the actual impact of each OPIC intervention, not on associating OPIC with the largest volume of private activity, or on insisting dogmatically that it stay out of almost all fossil fuels.