Risk, Uncertainty, and the Challenge of Sustainable Innovation

Innovation is risky business. For companies pursuing sustainable innovations, these risks take on the scale of the effort and the context of the problems, the politics, and the markets involved. The most important aspect of this challenge to sustainable innovation is understanding the nature of risk at work. Without this understanding, innovation efforts are paralyzed and innovation policies—especially those intending to promote new investments—stifle them instead.

(this post is part of a series attempting to recognize the unique challenges of sustainable innovation as I explore the implications for companes organizing for and managing such innovations)

Risk and Uncertainty

Everyone agrees innovation is risky business but few look past this offhand comment. Every new undertaking involves risk but, more importantly, every new undertaking is shaped by the nature of the risks involved.

Research clearly shows everyone from undergraduate pyschology students to psychology professors to corporate leaders to venture capitalists change their behaviors in the face of risk (see Daniel Kahnemann’s new book Thinking Fast and Slow for a great review of the literature by one of the founders of the field).

Unfortunately, while groundbreaking, most of this research has focused on showing what bad statisticians (or gamblers) we are—incapable of recognizing when a sure thing (e.g., an $8 gift) is worth less than an 80% chance at $11, and willing to risk more to avoid a loss than capture an equivalent gain. However, I have not been able to find in all of this research the recognition that not all risk is created equal.

That recognition comes from Frank Knight, an economist at the University of Chicago, who in 1921 showed how the term ‘risk’ actually refers to two very different types: ‘risk proper’ and uncertainty. ‘Risk proper’ is real; it can’t be eliminated but, because its probability can be accounted for with reasonable accuracy, it can be profitably managed and hedged against. Uncertainty, on the other hand, describes when the ‘risk proper,’ or probability, of a given outcome occurring is unknown. And not just unknown but when even the costs and benefits—the actual size of the investment required, the ultimate payoff, even the rules of the game—may be unknown and may change over time depending on the actions of others.

To Knight,

Uncertainty must be taken in a sense radically distinct from the familiar notion of risk, from which it has never been properly separated…. The essential fact is that risk means in some cases a quantity susceptible of measurement, while at other times it is something distinctly not of this character; and there are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating. It will appear that a measurable uncertainty, or risk proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all [italics mine].

We live with risk and uncertainty all the time. Anyone undertaking a new venture consciously or unconsciously weighs the likely benefits of success and costs of failure: whether to chase after a bus or not; to put 4 or 5 quarters in the meter or not, or to get in this line or that one at the grocery store. In each of these decisions, we weigh the costs and likely outcome (BTW, I should have gone with 5 quarters, trying so hard not to lose that last quarter cost me $40).

Nothing has more influence over the outcome of a new venture than this calculation—because it determines our very commitment to undertaking it. But risk proper and uncertainty have very different effects on our decisions.

Responding to uncertainty

The big difference between risk and uncertainty lies in their effect on our behavior. We can deal with risk proper. We can’t deal with uncertainty. We calculate the risks that can be calculated, make sure we can afford the losses (or offset them by hedging) and then proceed. In the face of uncertainty, however, we more often choose inaction.

This was Knight’s contribution: The reason more people don't innovate is not from an aversion to risk, but rather an aversion to uncertainty. Conversely, those who do undertake a new venture are not more risky—they just know something that others don’t. And that knowledge makes them less uncertain about the path forward.

Companies can offset the risks associated with investing in new manufacturing facilities or processing plants by attracting outside investors, taking out insurance policies, or capturing subsidies or loan guarantees to hedge against the (known) negative outcomes. But they can’t hedge against uncertainty—the odds, and the costs, and benefits, are unknown.

This isn't a new problem. Odysseus ran into it himself trying to get home, when he had to pass between the 6-headed Hydra, Scylla, and the ship-eating vortex Charybdis. Scylla lived in the cliffs on one side, and when ships came near would snap down and take 6 of their crew. Ships avoiding Scylla faced Charybdis, who might or might not rear up to take the entire ship. As the goddess Circe warned: “No, keep closer to Skylla's cliff, and row past that as quickly as may be; far better to lose six men and keep your ship than to lose your men one and all.” Like Odysseus, companies (and individuals) are always drawn to the known rather than the unknown risks.

Uncertainty, rather than risk proper is the predominant characteristic of the latter stages of innovation, when the stakes grow exponentially and missteps can take not only projects but also entire companies. Sure, early-stage R&D and market research can be costly, but represents only a small fraction of the cost necessary to bring innovations to market: the commercial-scale prototyping and demonstration of pilot plants and processes.

This is true for all new ventures but especially so for sustainable innovation. Take the cost to develop carbon capture and sequestration technology. Developing the technology and testing it in the lab may run in the tens of millions of dollars; building a demonstration plant costs around $700 million. The same holds for the development of manufacturing capacity, and the proof of scalability (and reliability) of any new products—as witnessed by the challenges to companies like Solyndra and Tesla. The larger the capital investments required to achieve scale, reliability, and profitability, the more salient the uncertainties become.

Sources of uncertainty in sustainable innovation

For the companies I spoke with during the research for the report, “The Business of Innovating: Bringing Low-Carbon Solutions to Market,” funded by the Pew Center on Global Climate Change, the most significant challenges in pursuing low-carbon innovation were the uncertainties surrounding public policies and market acceptance, followed by costs concerns and—only then—by technical uncertainty. As a United Technologies executive explained, “The potential and perceived value of new energy technologies can change quickly, and is significantly impacted by domestic and global public policy.”

In fact, the majority of corporate executives surveyed for that study believed that low-carbon innovation was indeed different from other types of innovation; 60% noting dependence on government policy as the most important difference. Indeed, 65% rated government policy the most significant uncertainty associated with low-carbon innovation (followed by 25 percent rating market uncertainty the highest).

I offer some brief descriptions of those sources of uncertainty here:

1. Policy Uncertainty Uncertainty surrounds which public policies will be enacted (or redacted), for how long, how much, and how they will be implemented. Such policies can take many forms, from emissions curbs to tax credits or subsidies. For example, the DOE’s loan guarantees—$465 million to electric automaker Tesla or $535 million to thin-film solar venture Solyndra—were meant to reduce the risks of lending to these particular firms but, in the process, created even greater uncertainty for competitors and investors in the same markets. Some incentives may be predictable and provide greater certainty; others may be unpredictable and increase uncertainty. Policy also acts more indirectly, by shaping the availability of raw materials for manufacture and use, shifting the regulatory environment of customers or suppliers, or supporting (or tolerating) nationally subsidized industries.

2. Market Uncertainty The second major source of uncertainty is market acceptance. Customer preferences regarding climate change shift across market segments and across time as competing alternatives emerge from within or outside an industry. And purchase decisions are driven by a customer’s own uncertainty about the direction of public policy. For suppliers to the automobile market, for example, the customers are the automakers—whose decisions hinge on upcoming emissions standards and other policies (and particularly how they will be measured). The same holds for energy generating equipment and the policy uncertainties that utilities, as customers driven by their local utility commissions, are expecting to face over the life of the plant. Similarly, building owners considering adopting energy efficiency innovations face a range of uncertainties regarding the technical and financial outcomes of such investments.

3. Business Uncertainty One of the largest of the business uncertainties associated with sustainable innovation is profitability. Many such innvoations attempt to displace well-established, well-entrenched competitors with mature supply chains and low-costs (brownfield rather than greenfield markets). While energy, food, and commodity markets are large, they are also very mature and offer little margin for error when deciding whether or not to enter them with an innovation. A miscalculated or unexpected costs or a reliability or performance shortcoming can spell the difference between a large profit and a large loss. Even a changed policy enforcement can turn what looked like a good business into a bad one. As one survey respondent stated, “Profitability is [the] main metric for all [low-carbon innovation] projects, yet sustainability issues are constantly changing the rules for [that] calculation. Long-term investments are often needed.”

4. Technical Uncertainty Finally, technical uncertainty contributes as well to the reluctance of companies to move beyond the R&D stage of innovations. From the cases I studied, two sources of technical uncertainty seemed the most salient. First was the uncertainty around whether new processes or new materials would work at commercial scale in the same ways they worked in the lab. This meant both "can we build a million units as reliably in a plant in Asia as we can build them in our lab?" and "will these components perform as expected for the 40-year life of the plant?" The second source of uncertainty had to do with the interdependent and long-lived nature of most innovations targeting brownfields markets: the need to ensure that everyone from suppliers to regulators to even competitors agrees to follow the same technical standards. The sheer size of many of these markets requires that everyone move in the same direction or not at all.

Facing long-term and large commitments to scaling up sustainable innovation, these uncertainties have a significant effect on how and when companies will pursue opportunities. With the ability to identify and manage or resolve policy and market uncertainties, companies can act on growing opportunities sooner and with more commitment; without that ability, companies often adopt a wait-and-see posture that follows, rather than leads, changes in their industries.

As Knight argued, it is critical to understand when decisions are driven by risk and when they are driven by uncertainty as “[t]here are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating.” Policies that subsidize risk do not necessarily remove uncertainty and, indeed, can create greater uncertainty by both their presence and by the added possibility of their removal.

Comments

[BrianSJ3]

In Hacking in 'The Emergence of Probability' points out that we could have had two quite separate words for epistemic and aleatoric uncertainty, but happened to end up with the two concepts conflated.

[BrianSJ3]

Typo - sorry Ian Hacking, not In Hacking!