Standard competitive theory rests on a heroic assumption: that buyers and sellers share symmetric knowledge about the goods being exchanged. Relax this assumption, and the elegant conclusions of the First Welfare Theorem begin to unravel. Even in markets with many sellers, free entry, and homogeneous underlying technologies, informational disparities can generate substantial and durable market power.

The theoretical foundations trace back to Akerlof's lemons problem, Rothschild-Stiglitz screening models, and Spence's signaling framework. Yet the implications extend far beyond adverse selection. When one side of the market possesses superior information about product quality, service necessity, or transaction consequences, that informational advantage translates directly into pricing discretion and rent extraction.

This article examines three interconnected phenomena: how expertise itself functions as an entry barrier, how credence goods generate systematic distortions in provider incentives, and how regulatory instruments attempt to restore something approximating competitive discipline. Throughout, the analytical challenge is distinguishing informational rents from returns to genuine investment in human capital—a distinction with substantial normative weight for policy design.

In markets for professional services—medicine, law, financial advising, technical consulting—the very knowledge that qualifies a provider to deliver value simultaneously insulates them from competitive pressure. Consumers cannot meaningfully evaluate quality before purchase, and often cannot evaluate it after purchase either. This asymmetry is not incidental; it is constitutive of the transaction.

Formally, when the marginal buyer's assessment of provider quality has high variance relative to true quality dispersion, price competition becomes attenuated. Providers can sustain markups above marginal cost without triggering entry, because entrants face the same evaluation problem magnified: they lack established reputational signals that would allow buyers to distinguish them from lower-quality alternatives.

This creates what mechanism designers call endogenous market power—rents that arise not from artificial restrictions but from the informational structure of the exchange itself. Klein and Leffler's model of quality-assuring price premia formalizes this: prices must exceed competitive levels sufficiently to make quality cheating unprofitable, generating persistent supracompetitive margins as an equilibrium requirement.

The empirical signature is distinctive. Professional service markets exhibit price dispersion far exceeding cost dispersion, weak correlation between price and observable quality, and substantial geographic variation unexplained by input costs. Entry occurs, but slowly, and typically requires costly reputation-building investments that themselves function as commitment devices.

Critically, this market power is not eliminated by increasing the number of providers. Adding physicians to a local market may reduce prices modestly, but the informational structure ensures that residual rents persist. The competitive benchmark itself becomes the wrong analytical frame.

Takeaway

Market power can emerge from the structure of knowledge itself, not merely from scarcity or regulation. When buyers cannot verify what they are purchasing, competition disciplines prices only weakly.

Credence goods—following Darby and Karni's original formulation—are those where the seller both diagnoses the buyer's need and supplies the remedy. Auto mechanics, dentists, IT consultants, and financial advisors all operate in this domain. The provider's dual role as diagnostician and supplier creates fundamental incentive incompatibilities that no amount of competition can resolve.

Dulleck and Kerschbamer's canonical framework identifies three distinct pathologies: overtreatment (recommending unnecessary services), undertreatment (providing insufficient care while charging for more), and overcharging (billing for services beyond those delivered). Each pathology responds differently to institutional structure.

Verifiability and liability shape which distortions dominate. When treatments are verifiable but diagnoses are not, overtreatment tends to prevail—providers recommend expensive interventions the buyer cannot dispute the necessity of. When neither is verifiable, undertreatment becomes tempting, since providers can charge for premium services while delivering minimal ones.

Reputation mechanisms provide partial discipline, but only under restrictive conditions. Repeat interactions must be frequent, information must diffuse across the buyer population, and the discount rate must be sufficiently low. For infrequent, high-stakes purchases—major medical procedures, complex litigation—these conditions typically fail, leaving buyers structurally vulnerable.

The welfare consequences are subtle. Efficient production of credence goods requires that providers exercise costly diagnostic effort. Any mechanism that eliminates provider rents entirely destroys this incentive, generating a fundamental tradeoff between rent extraction and diagnostic accuracy.

Takeaway

When the person who tells you what you need is the same person who sells it to you, competitive markets do not automatically produce efficient outcomes—regardless of how many sellers exist.

Three regulatory approaches dominate policy responses to information-based market power. Licensing regimes restrict entry to qualified providers, ostensibly ensuring baseline competence. Disclosure requirements force information transmission from informed to uninformed parties. Liability rules reallocate the cost of poor outcomes to induce provider care.

Each instrument carries characteristic failure modes. Licensing generates rents for incumbents and often expands beyond genuine quality assurance into pure entry restriction—Kleiner's extensive empirical work documents licensing's frequent capture by professional associations pursuing supracompetitive returns rather than consumer welfare.

Disclosure regimes suffer from what Ben-Shahar and Schneider term the failure of mandated disclosure: information provided is rarely processed, comprehension gaps between provider and consumer persist despite formal disclosure, and disclosures often serve as liability shields rather than genuine informational remedies.

Liability rules, particularly negligence-based standards, can align incentives more effectively but require verifiable outcome measures. In credence goods markets where outcomes depend on unobservable inputs and stochastic factors, distinguishing negligent from unlucky outcomes becomes prohibitively costly, undermining the mechanism's disciplinary function.

The mechanism design perspective suggests hybrid approaches: outcome-contingent payments combined with limited entry restrictions and targeted disclosure of decision-relevant rather than comprehensive information. Kessler and McClellan's work on hospital reimbursement demonstrates how such combinations can reduce distortion without eliminating the rents necessary to sustain quality provision.

Takeaway

Every regulatory response to information asymmetry involves tradeoffs between rent extraction, provider incentives, and administrative feasibility. There is no clean solution—only better and worse mixtures of imperfect instruments.

Information asymmetry does more than distort otherwise competitive markets—it can render the competitive benchmark itself analytically inappropriate. When knowledge distribution is fundamentally uneven, market power emerges endogenously and persists despite structural conditions that would otherwise support competition.

For policy design, this reframes the central question. Rather than asking how to make credence goods markets competitive, we should ask how to structure institutional arrangements that align provider incentives while preserving the rents necessary to sustain expertise investment. The mechanism design literature offers tools for this analysis, but implementation requires accepting that some informational rents are efficiency-enhancing rather than welfare-reducing.

The practical implication for market regulation is humility about first-best benchmarks and precision about which distortions matter most in particular contexts. Second-best analysis, properly conducted, is not a retreat from rigor—it is where rigor becomes genuinely useful.