Trends In a Time of Chaos
In 2010, solar energy cost $0.35 per kilowatt-hour to produce. This price had been inching down for decades, but very slowly. Based on this past performance, the International Energy Agency was able to confidently predict that by 2022, the price of solar power would be roughly $0.21/KWh. When actual solar prices began to plunge, they repeated their analysis in 2014, and revised their prediction to about $0.12/KWh. [https://thenarwhal.ca/renewable-energy-growth-again-blows-eia-forecasts-out-water/]
As of January, 2022, the unsubsidized cost of solar electricity in the US was about $.06/KWh. [https://www.statista.com/statistics/493797/estimated-levelized-cost-of-energy-generation-in-the-us-by-technology/] The IEA’s 2010-2014 forecasts were not just wrong, even their revised estimate was off by a factor of two. This failed projection has become an object lesson in the pitfalls of trends analysis that is based solely on past performance metrics. In short: the IEA looked at the numbers while ignoring the real-world buzz of invention and investment going on all around them. Their analytical tools were uncoupled from reality.
Trends are changes in sequential data gathered over time. Trend analysis is a very powerful tool, without which critically important initiatives such as the Intergovernmental Panel on Climate Change would not be possible. Properly used, trend analysis is extremely useful. Executed poorly, or used incorrectly, it can be little more than bullshit.
The failure of the IEA solar projections points to an equally important shortcoming of the method. Its very strength—that it reduces complexity of analysis to one or a few numbers that can be tracked—is also its weakness. We find the same Achilles Heel in many if not most approaches to foresight and innovation research. Researchers try to reduce the complexity of the analytical process as much as possible; to make it inexpensive; and to make it usable by as many people as possible. These criteria push the process in the direction of simplification and abstraction—away from the individual cases and the experience of individual specialist researchers, and in the direction of a Taylorist world of interchangeable workers and units of work. Also, the more a set of shifting data can be abstracted numerically, the more its analysis can be computerized. These are all real advantages in a highly competitive business environment.
If there were no pressures to produce quarterly reports, to keep up with the competition, and to make everything easy and digestible, would researchers make these kinds of abstractions? Given all the time in the world to do the work, they might but only in addition to a very different set of analyses—a more deliberately human-centric set.
In 2010 there were analysts following the trends of solar energy who were also keenly interested in the specific people doing the work and in fringe projects that had no guarantee of paying off. There were a lot of these—in fact, more and more as momentum slowly grew in the renewables sector. The indicators of change were not to be found in the metrics that had previously been used to track the trends; they were in the mounting reports of breakthroughs and the increasing excitement of researchers who were finally cracking the hard problems of commercializing Solar. Researchers who took a radically human-centric approach—getting down in the trenches and talking to people, reading all the papers as they were published, and understanding the community that was coalescing around renewable energy—were much better positioned to understand what was and was not relevant in the trends metrics.
Several sectors are currently poised to change rapidly in the same way that Solar has over the past decade. A good example is batteries and power storage in general. The price of batteries has declined by 97% in the past three decades—but, dramatic as this trend has been, it may not capture the actual rate of change of the next few years. This is because there is now a large and growing community dedicated to solving energy storage problems. The pace of discoveries continues to increase, with new materials and entirely new paradigms such as gravity storage, in which cranes or rail cars on steep hillsides use wind or solar power to lift massive weights which then produce power as they reverse the winching process. Such a system, which doesn’t even use batteries and so completely sidesteps all the issues associated with them, is the kind of signal that will not be reflected in, eg., trends in the price of lithium. The way to know about its significance is to be involved in the sector and understand how its culture is evolving. Another sector that is breaking its trendline is Space, where one company alone, SpaceX, in the past several years launched as much mass into orbit as every governmental entity has since Sputnik. In 2022, SpaceX’s Superheavy launcher has not flown yet and no one knows whether it even will. A trends analysis based on existing launch capabilities would not take Superheavy into account; yet even if it doesn’t work, similar launchers such as Rocket Labs’ Neutron are being developed. These new systems represent a step change in launch capability and may create new markets. Thomas Watson, chairman of IBM, famously said in 1943, “I think there is a world market for maybe five computers.” Based on trends in launch costs to date, we could make a similar claim about the market for orbital, lunar, and asteroid missions.
The point is not that trends analysis is not a powerful and highly accurate tool, but that there is a point at which it is no longer useful to abstract away from real people and all the messy complexity of changing communities of innovators and investors. Here, a different approach to research has to come into play: the approach of radical human-centricity. This research mode does not abandon trends, but always digs into the data behind them to explore the communities that are behind the abstractions. By taking an ethnographic, personally involved approach to understanding these communities, researchers and innovators can see past the data points to the real causes of present change—and use that knowledge to achieve commercial success.