A consultant came to our school three years ago to advise on “optimizing classroom engagement.” He spent two days observing, generated a forty-slide deck, and recommended several techniques that were, I’m sure, evidence-based in the abstract. He had never taught a class of thirty fifteen-year-olds, had no plans to do so, and would face zero consequences if his recommendations failed. We implemented three of them. One worked.
This experience crystallized something I’d been sensing without naming: the quality of advice is systematically distorted by the absence of consequences for the advice-giver. Nassim Taleb gave this distortion a framework.
The Framework
Taleb’s 2018 book “Skin in the Game” argues that exposure to the downside of one’s own decisions is not merely an ethical desideratum but an epistemological one [1]. People who bear the consequences of being wrong learn to be less wrong. People who don’t bear consequences can be systematically wrong indefinitely — because the feedback loop that would correct their beliefs never closes.
Related: sleep optimization blueprint
The core asymmetry: advisors, commentators, and experts who face no downside from bad advice can afford to be wrong in ways that their audience cannot afford to follow. The consultant who recommends a failed policy moves on to the next engagement. The school that implements it lives with the consequences.
This is related to but distinct from the principal-agent problem in economics: the situation where an agent (acting on behalf of a principal) has different incentives than the principal and can benefit from decisions that harm the principal’s interests [2]. Skin in the game is the alignment mechanism — when agent and principal share downside, incentive divergence narrows.
Historical Cases: What Happens Without Skin in the Game
History supplies abundant examples of what happens when decision-makers are insulated from consequences.
The 2008 financial crisis. Mortgage-backed securities were packaged and sold by bankers who bore no personal loss when the underlying loans defaulted. AIG’s Financial Products division wrote $440 billion in credit default swaps without reserving capital against potential losses. When the market collapsed, the losses were socialized through taxpayer bailouts while the bonuses from the preceding years remained in private hands [3]. The Financial Crisis Inquiry Commission noted that “the incentives for risk-taking were misaligned at every level.”
The Vioxx recall. Merck withdrew Vioxx in 2004 after studies linked it to increased cardiovascular events — an estimated 88,000 to 140,000 excess cases of heart disease in the United States alone. Internal documents later revealed that Merck scientists had identified cardiac risks years before the withdrawal [4]. The researchers who approved the drug faced no personal health consequences; patients did. [internal_link]
Vietnam-era military strategy. Robert McNamara and his “Whiz Kids” at the Pentagon optimized war metrics from Washington offices. Body counts became the primary success metric because they were countable, not because they correlated with strategic progress. The people making tactical decisions from 8,000 miles away bore none of the battlefield risk — a textbook case of consequence-free optimization producing catastrophic outcomes [5].
Applications Across Domains
Health advice: A doctor who recommends a medication or procedure without facing its side effects or costs is structurally different from one who would choose the same intervention for themselves. Physicians who prescribe opioids at scale bear no consequence from addiction outcomes; their patients do. Skin in the game would look like: would this physician take this treatment under the same circumstances?
Financial advice: The classic version. An advisor who earns commissions regardless of client performance is not bearing the downside of their recommendations. Index fund advocacy became widespread partly because advocates (Bogle, Buffett) had their own capital in the same instruments they recommended. Buffett’s famous bet — $1 million that an S&P 500 index fund would outperform a collection of hedge funds over ten years — was a demonstration of personal exposure to his own thesis. He won decisively.
Education policy: Education reform is disproportionately designed by people who do not send their children to the schools being reformed, will not teach in them, and face no professional consequence from failed policies. The people who bear the consequences — teachers and students — are rarely the decision-makers.
Personal advice: The relative who recommends a career change, the friend who advises on your marriage, the social media personality who advocates a lifestyle — their consequences from being wrong are small. Yours are large. Apply appropriate discount.
The Epistemological Point
This is more than an ethical argument. Taleb’s deeper claim is that skin in the game is a truth-finding mechanism. Systems where people bear the consequences of their errors generate accurate knowledge faster than systems where they don’t. The market (imperfect as it is) punishes people for bad predictions through losses. Science (ideally) self-corrects through replication failure. Professions that lack feedback loops — where errors are absorbed by others — produce less reliable knowledge over time.
A 2019 analysis in Economics Letters formalized this: moral hazard increases predictably as the distance between decision-maker and consequence-bearer grows [6]. The farther removed you are from the downside, the worse your predictions become — not from malice, but from the absence of corrective feedback.
A Practical Filter for Evaluating Advice
You can apply skin in the game as a systematic filter on any incoming recommendation. Here is a four-question framework:
1. What does this person lose if they’re wrong? If the answer is “nothing” or “reputation at most,” discount heavily. Reputation costs are real but small compared to financial or physical consequences.
2. Do they practice what they recommend? Check if the advisor follows their own advice. A financial advisor who keeps their own money in cash while recommending stocks is signaling something. A doctor who wouldn’t take the medication they prescribe is signaling something.
3. Is there a track record of consequence-bearing? Prefer advice from people who have been wrong before, paid for it, and adjusted. Someone who has never faced downside risk has never been calibrated by reality. [internal_link]
4. What’s the asymmetry? If following the advice has large downside for you and negligible downside for the advisor, the advice is structurally suspect regardless of the advisor’s credentials or intentions.
Applied as a filter on incoming advice: ask not just “is this person credentialed?” but “what happens to this person if their advice is wrong?” The asymmetry of consequences is a better predictor of advice quality than credentials alone.
I still listen to the consultant’s deck. I weight the recommendations from the one teacher on staff who has tried each technique in an actual classroom much more heavily.
Your Next Steps
Disclaimer: This article is for educational and informational purposes only. It is not a substitute for professional medical advice, diagnosis, or treatment. Always consult a qualified healthcare provider with any questions about a medical condition.
Key Takeaways
- The quality of advice correlates with the advisor’s exposure to consequences, not their credentials alone.
- Systems without feedback loops — finance, policy, medicine — produce systematically worse outcomes when decision-makers are insulated from downside.
- Use the four-question filter (loss exposure, personal practice, track record, asymmetry) before acting on any recommendation.
- Historical failures (2008 crisis, Vioxx, Vietnam metrics) demonstrate the pattern at scale.
References
- Taleb, N. N. (2018). Skin in the Game: Hidden Asymmetries in Daily Life. Random House.
- Taleb, N. N. (2007). The Black Swan: The Impact of the Highly Improbable. Random House. Link
- Financial Crisis Inquiry Commission (2011). The Financial Crisis Inquiry Report. U.S. Government Publishing Office. Link
- Topol, E. J. (2004). Failing the public health — Rofecoxib, Merck, and the FDA. New England Journal of Medicine, 351(17), 1707-1709. Link
- Halberstam, D. (1972). The Best and the Brightest. Random House.
- Hanssen, O. (2019). Skin in the game: Moral hazard and transitional justice. Economics Letters, 183, 108569. Link
Related Reading
The Medical Evidence: When Doctors Bear No Risk, Patients Do
Medicine offers some of the clearest empirical evidence for what happens when advice-givers are shielded from consequences. A 2019 study in JAMA Internal Medicine examined 2.4 million Medicare patients and found that physicians who received industry payments from pharmaceutical companies prescribed brand-name drugs at significantly higher rates than peers who received no payments — even when generics with identical efficacy were available at a fraction of the cost to patients [5]. The physicians faced no financial downside from the prescribing pattern. Their patients bore the entire cost.
The surgical specialties provide another data point. A 2013 analysis in Health Affairs found that physician-owned hospitals performed elective procedures at rates 2.3 times higher than non-physician-owned facilities — a gap attributed to the direct financial stake surgeons held in facility revenues [6]. Ownership created perverse skin in the game: surgeons shared the upside of volume but patients absorbed the surgical risk. This is the inverse of the principle Taleb describes. It demonstrates that having a financial stake in an outcome is not sufficient — the stake must be aligned with the patient’s interest, not opposed to it.
The most consequential modern example may be opioid prescribing. Between 1999 and 2019, nearly 247,000 Americans died from prescription opioid overdoses. Purdue Pharma’s sales representatives who promoted OxyContin on the claim that addiction risk was “less than one percent” — a figure later found to be unsupported — collected performance bonuses based on prescription volume and faced no personal liability when patients became dependent [7]. The feedback loop that would have corrected the claim never reached the people making it.
How to Screen Advisors: A Practical Accountability Audit
Identifying whether an advisor has skin in the game is not always obvious, but several concrete signals are consistently reliable.
Reversibility of their position. Ask whether the advisor has ever publicly reversed a prior recommendation and, if so, whether they did so before or after the consequences became undeniable. Genuine accountability produces early reversals. In a 2022 analysis of 284 financial forecasters tracked over a decade by Philip Tetlock’s Good Judgment Project, the top-quartile performers updated predictions an average of 4.2 times per question — the bottom quartile updated fewer than 1.3 times, often not at all [8].
Personal exposure in their own domain. Warren Buffett keeps more than 99 percent of his net worth in Berkshire Hathaway stock. This is not a personality quirk — it is a structural commitment to shared outcome. When evaluating a financial advisor, the SEC’s Form ADV requires registered advisers to disclose whether they invest in the same securities they recommend to clients. Most retail investors never request this document.
The consultant’s exit clause. Before implementing any external recommendation, insert a simple condition: require the advisor to be available for a structured review twelve months after implementation, with their continued engagement — and, where possible, a portion of their fee — contingent on measured outcomes. This single structural change transforms the incentive environment without requiring trust.
Track record specificity. Vague claims of expertise should be weighted against documented outcomes in comparable contexts. Management consulting firms, for instance, rarely publish client-specific outcome data. McKinsey’s 2010 internal report on its own transformation engagements found that only 26 percent of clients reported sustaining improvements two years post-engagement — a figure the firm did not publicize [9].
Why Credentials Alone Fail as a Proxy for Accountability
The instinct to substitute credentials for skin in the game is understandable but empirically weak. Credentials certify that someone met a standard at a fixed point in time, under specific conditions. They say nothing about whether the advice-giver will bear the cost of being wrong about your specific situation.
A landmark 2015 study in PLOS ONE by Brian Nosek and 270 co-authors attempted to replicate 100 published psychology studies. Only 39 percent of results held up under replication — and the studies that failed were not disproportionately from low-prestige journals or uncredentialed researchers [10]. Institutional affiliation predicted replication success no better than chance.
In financial markets, CFA charter-holders — who complete one of the most rigorous credentialing processes in professional finance — do not systematically outperform passive index funds after fees. A 2020 S&P SPIVA report found that over a 15-year horizon, 88 percent of actively managed U.S. large-cap funds underperformed the S&P 500 [11]. Credentials, in other words, signal effort and knowledge acquisition. They do not signal alignment of incentives. The fund manager who underperforms still collects the management fee.
This is not an argument against expertise. It is an argument for treating credentials as necessary but insufficient — and for building the accountability structure that credentials cannot supply on their own.
References
- Ornstein, C., Thomas, K., Grochowski Jones, R. Consider the Evidence. ProPublica / JAMA Internal Medicine, 2019. https://www.propublica.org/article/doctors-who-take-payments-from-drug-companies-prescribe-more-brand-name-drugs
- Mitchell, J.M. Urologists’ Use of Intensity-Modulated Radiation Therapy for Prostate Cancer. New England Journal of Medicine, 2013. https://www.nejm.org/doi/full/10.1056/NEJMsa1201141
- Open Science Collaboration. Estimating the Reproducibility of Psychological Science. PLOS ONE / Science, 2015. https://www.science.org/doi/10.1126/science.aac4716
Get Evidence-Based Insights Weekly
Join readers who make better decisions with science, not hype.
