Humans are naturally optimistic. That might sound untrue, and yet it’s well-documented in psychology and, as is the focus of this post, in public sector commissioning.
Take this high-information world in which we live. You often hear phrases such as “it’s just all so doom and gloom” when discussing the news, social media or current events. Generally, people attribute that doom and gloom to the information reaching our phones and TVs – war has broken out again, disease is spreading, the latest economic growth figures are released and on and on it goes. If only we lived in a low-information world we could live in ignorant bliss. And this is precisely the underpinning of optimism bias – far from perfectly rational economic agents we are, as a species, rather prone to assuming the best of things when we have imperfect information. Reality is often not quite so rosy.
The theory behind optimism bias is simply that when we do not have quality information about something, we tend to overestimate the positives and opportunities and downplay the negatives and risks. In economic studies, when examining the merits of, most commonly, government projects that will cost taxpayers huge sums of money, that optimism can mean poor project outcomes, overrunning and overspending.
So, how do we account for this and ensure we are providing strong, realistic estimates for commissioners?
Correcting optimism bias
As economists, we want to separate out the bias so we are left with realistic central estimates – easier said than done. Let’s start with some factors that optimism bias typically depends upon:
- Quality of underlying evidence – gaps in evidence leave room for optimism bias to grow
- Relevance of underlying evidence – underlying evidence might not reflect the conditions in which the intervention will be used in the future, such as if it is to be used in different settings or if imperfect implementations are to occur
- Use of assumptions – assumptions might be made, sometimes implicitly, with high uncertainty
When conducting a study, a number of sources of bias can be eliminated in the design. That is, however, not always possible when the resources are not available to conduct new or comprehensive studies.
Where these factors can’t be corrected, we must have a framework to deduce the appropriate optimism bias correction. We have two things we can rely upon – one, we can grade the confidence that we have in our estimates according to the presence of the limitations illustrated above. Secondly, we are fortunate to be able to rely upon guidance produced by HM Treasury (the ‘Green Book’) which provides estimates, grounded in historical empirical data, on optimism bias ranges for different types of projects across the public sector.
Why should we correct for optimism bias?
When conducting economic analyses, we have to think about the purpose of the analysis that we are conducting. At Unity Insights, we produce our economic analyses to support commissioners across health and social care to make investment decisions, which also helps innovators as the results are aimed towards the audience that they need to convince. In essence, we help to build the business and the economic case for investment.
Public sector commissioning demands value for money. As a result, commissioners want to ensure that they invest only in projects that have a high probability of generating a positive return on investment. The most straightforward way to ensure this is to be prudent with your estimates – when uncertain, err on the side of underestimating the benefits while being particularly aware of the upside risk to costs. If a project or innovation still generates a positive return on investment with a prudent approach, you can have much greater certainty in the case for investment. Optimism bias is a particularly useful tool to ensure prudency, build a stronger business case, and give confidence to commissioners.
The public sector focused approach differs from the purposes of academia, for example, where understanding the true intervention effect is key. Optimism bias is also reduced by practitioners when the evidence is stronger through more rigorous real-world data.
Sensitivity analysis and costed risks are other methods to manage uncertainty in our economic estimates, but the former does not correct for bias (merely demonstrating the distribution of uncertainty on either side of our central estimates) and the latter often relies upon data which might not be readily available.
Summary
To summarise, tackling optimism bias in economic analyses is crucial for building trust and credibility with stakeholders. By carefully addressing this bias, we make our forecasts more reliable and give decision-makers realistic, practical insights. This thorough approach helps the public sector, including the NHS, to make wise investment choices, protect public funds, and maximize benefits for society. At Unity Insights, our dedication to delivering accurate, bias-free analyses is central to our goal of driving meaningful, value-driven change in health and social care.