Job security isn't what makes people take risks. In this study of inventors, it made their work safer.

You have a strong performer you want to keep, and the obvious tool is a promise: a protected role, a long runway, an assurance that one miss will not end the conversation. The assumption behind that promise is that people who feel safe will reach further. A paper in Organization Science, published in October 2026, tests that assumption on corporate inventors, and the answer it reports is not the comfortable one.

The paper is "The Paradox of Inventors' Job Security," by David Wehrheim of IESE Business School, University of Navarra, and Eduardo Melero and Neus Palomeras, both of the Universidad Carlos III de Madrid. One note on how far I got: the publisher blocked me from the full text, so everything below comes from the published abstract, which I read through Crossref's record. I have not seen the tables, the effect sizes or any limitations section, and I will not describe them.

Here is what the abstract says they did. They built matched employer–employee data from patent records, then used the staggered adoption of wrongful discharge laws across U.S. states between 1980 and 1999 as a source of change in job security. Using cross-state variation over time in a difference-in-differences framework, they compared inventors whose states strengthened protection with inventors whose states had not yet done so. This is archival data on patents, not a survey and not an experiment, though the timing of the laws is what lets the authors argue about cause.

The finding, in the abstract's words, is that increased job security leads inventors "to produce innovations that rely more intensely on their firms' technologies, are less novel, and generate higher private value to their employers." The authors' explanation is about incentives:

Thus, whereas employment protection provides inventors with a safety net to fail that could facilitate exploration, it instead drives them to produce innovations that draw more heavily on internal technologies and are less novel.Wehrheim, Melero and Palomeras, abstract of "The Paradox of Inventors' Job Security," Organization Science, as carried by Crossref

Read the sentence slowly. A safety net could have encouraged exploration, and the authors say so. What they report is that the effect that dominated was a different one: secure people expect to stay, so they invest in knowledge that is useful mainly inside the company they expect to stay at. That knowledge is firm-specific. It makes their next piece of work better for you and less novel for everyone else.

Note the third item in the finding, because it complicates the easy reading. The work was not worse for the employer. The abstract says it generated higher private value to the employers. So the study does not show that security damages results. It shows that security changes what kind of results you get: more exploitation of what the firm already owns, less reach beyond it.

Here is why this lands differently for the person at the top than for anyone else in the building. A manager can grant a protected role. Only the person at the top decides what the organization is trying to buy with it. If the goal is depth in your own technology, the study's pattern is a fair description of a payoff. If the goal is novelty, the same protection may be working against you, and nobody beneath you has the standing to point that out. The people who benefit from the guarantee are not going to report that it narrowed their ambition, and they may not notice it themselves.

The part the research doesn't say, but that I'll say plainly: this study measured legal employment protection across states, applied to inventors who file patents. It did not measure a manager's private promise, or a company's own policy, or anyone outside technical invention. Whether an informal assurance from you behaves like a state statute is my inference, and it may not hold. The mechanism the authors propose, expecting to stay and investing accordingly, is at least plausible for a promise that people actually believe. That caution is mine, not the authors'.

I also cannot tell you from the abstract how large the shifts were, or whether they differed across kinds of firms or inventors. Those are the questions I would want answered before changing any policy.

What you can do this week: list the people in roles where you have, formally or by habit, removed the risk of failure. Beside each name, write one sentence on what that role is supposed to produce for you, deeper work on what the company already knows or something it does not. Where the sentence says novelty, look at the last year of their output and ask whether it has stayed close to home. If it has, the safety may be doing what this paper describes. Then decide on purpose where you want the guarantee and where you want the exposure.

Seeing which of your commitments are quietly steering behavior is the kind of thing the Leadership Profile is built to surface. You can start from the wiring map.