·Economics & Markets
Section 1
The Core Idea
Developing a new pharmaceutical drug costs an average of $2.6 billion and takes twelve to fifteen years from discovery to FDA approval. That figure — established by the Tufts Center for the Study of Drug Development in 2014 — is not a measure of pharmaceutical complexity. It is a barrier to entry. It means the number of organizations on earth capable of bringing a new drug to market is measured in dozens, not thousands. The $2.6 billion pays for the clinical trials that 90% of drug candidates fail, the regulatory expertise that takes decades to accumulate, and the manufacturing infrastructure that costs billions before a single pill ships. Anyone can have an idea for a drug. Almost no one can afford to make it real.
Barriers to entry are the structural obstacles that prevent new competitors from entering an industry or market. The concept was formalized by economist Joe Bain in his 1956 work "Barriers to New
Competition," which identified three primary sources: economies of scale that give incumbents lower unit costs, product differentiation that creates customer loyalty to established brands, and absolute cost advantages from control of essential resources or proprietary technology. Michael Porter expanded the taxonomy in "Competitive Strategy" (1980) to include capital requirements, switching costs, access to distribution channels, and government policy. The list has grown. The insight hasn't changed: the height of entry barriers determines whether an industry's profits are temporary or structural. High barriers mean incumbents keep their margins. Low barriers mean competition erodes them to commodity levels. The airline industry illustrates the erosion side: despite generating over $800 billion in global revenue, airlines have historically earned returns at or below their cost of capital because entry barriers — while not trivial — have never been high enough to prevent new carriers from launching and undercutting fares.
The taxonomy matters because different barrier types have fundamentally different characteristics. Regulatory barriers — FDA drug approval, banking licenses, telecom spectrum auctions — are created by government and can only be changed by government. They tend to be the most durable because breaching them requires lobbying, not capital. Capital barriers — a semiconductor fab costs $20 billion, a new commercial aircraft program costs $15–25 billion — select for deep pockets but can theoretically be overcome by anyone willing to write the check.
Scale barriers operate through cost curves: TSMC's per-chip manufacturing cost at 90% utilization is structurally below any competitor running at 60%, and closing that gap requires matching TSMC's volume before matching its prices. Network barriers are the newest and potentially most powerful: Visa's 4.3 billion cards in circulation make the network indispensable to merchants, who make it indispensable to cardholders, who make it indispensable to merchants. Each barrier type creates a different kind of problem for would-be entrants, and the most protected industries stack multiple types simultaneously.
The counterintuitive dimension: barriers to entry constrain incumbents as much as they exclude challengers. The same $20 billion semiconductor fab that blocks new entrants also commits the incumbent to a multi-year investment cycle with no guarantee of return. The same FDA approval process that protects drug margins means pharmaceutical companies spend $2.6 billion before learning whether a product works. The same banking regulations that limit competition impose compliance costs consuming 6–10% of large banks' total expenses. Barriers are walls, and walls restrict movement in both directions. Industries with the highest barriers tend toward oligopoly — a handful of players locked together in a structure none can easily exit and none of their potential competitors can easily enter. The semiconductor industry consolidated from dozens of leading-edge manufacturers in the 1990s to effectively three — TSMC, Samsung, and Intel — precisely because capital barriers eliminated everyone else while locking the survivors into ever-increasing investment cycles.
The most protected industries don't rely on a single barrier type. They stack them. Standard Oil combined scale barriers (90% of refining capacity), distribution barriers (exclusive railroad rebates), capital barriers (vertically integrated supply chain), and regulatory barriers (political influence that delayed antitrust action for decades). The modern equivalent is the pharmaceutical industry: regulatory barriers (FDA approval takes twelve years), capital barriers ($2.6 billion per drug), intellectual property barriers (twenty-year patents), scale barriers (global clinical trial infrastructure), and distribution barriers (relationships with hospitals, pharmacies, and insurance companies). A challenger must breach all five layers simultaneously. Breaching four out of five is insufficient — each remaining layer can independently block entry.
The stacking principle explains why some industries attract constant disruption talk but no actual disruption. Banking has been "about to be disrupted by fintech" for over a decade. The barriers — regulatory capital requirements, compliance infrastructure, deposit insurance, and decades of consumer trust — have held because no single fintech innovation can breach all layers at once.
The relationship between barrier height and industry structure is one of the most empirically validated patterns in economics. Bain's original 1956 study found that industries with "very high" barriers — automobiles, cigarettes, typewriters, steel — earned profit rates roughly twice those of industries with "low" barriers. Subsequent research by Michael Porter, Richard Caves, and others confirmed the pattern across geographies and decades. The credit rating industry (three firms, 40%+ margins for decades), the commercial aircraft industry (two firms, persistent profitability despite cyclical demand), and the payment network industry (two dominant networks, 60%+ operating margins) all follow the same logic: high barriers produce concentrated structures that sustain above-average returns. Restaurants, hair salons, and landscaping businesses — industries with trivial barriers — produce millions of competitors and margins barely above subsistence. The barrier determines the structure. The structure determines the returns.
