International Industrial Economics Flashcards

Foundations of Industrial Organization and Antitrust Thinking

Industrial Organization, or IO, is the specialized study of how corporations compete for consumer capital in real-world environments. Departing from the theoretical ideal of "perfect competition" where all participants act equitably, Industrial Organization focuses on firms that possess market power and the strategic methods they employ to suppress rivals and maximize wealth extraction. To grasp the mechanics of IO, one must understand the perspective of Antitrust Authorities, whose viewpoints have evolved through three distinct historical eras. In the 1950s, the field was dominated by the Structure-Conduct-Performance (SCP) Paradigm, which posited that market structure dictates firm behavior. Under this view, a market with few large firms (Structure) would inevitably lead to collusion (Conduct) and consumer exploitation. Consequently, the prevailing regulatory philosophy was that "Big is Bad," leading authorities to break up large entities.

During the 1970s, the Chicago School of economics at the University of Chicago challenged this notion. They argued that a monopoly might arise simply because a firm is exceptionally efficient and superior to its competitors. In their view, punishing a firm for its size was effectively punishing its success. They advocated for a hands-off approach, trusting that the free market would eventually correct itself. Today, modern IO utilizes Game Theory to adopt a more nuanced stance. Recognizing that the Chicago School was perhaps too optimistic, economists acknowledge that firms do engage in predatory pricing and exclusive deals to obstruct new market entrants. Current regulatory practice does not assume that large size is inherently negative, but it maintains a skeptical watch, analyzing every market scenario on a case-by-case basis using strategic modeling.

Measuring Market Power: The Lerner Index, HHI, and the SSNIP Test

Market power is defined as the mathematical ability of a firm to raise its prices above its Marginal Cost ($MC$). The primary tool for measuring this markup is the Lerner Index ($L$), expressed by the formula:

$L = \frac{P - MC}{P} = \frac{1}{|\epsilon|}$

In this equation, $\epsilon$ represents the price elasticity of demand. If consumers have a desperate need for a product (inelastic demand), the firm can command a significant markup, and the Lerner Index approaches the value of $1$. Conversely, if consumers have many alternatives (elastic demand), markups become difficult to maintain, and the Lerner Index drops toward $0$. To assess the concentration of an entire market, regulators utilize the Herfindahl-Hirschman Index ($HHI$). This is calculated by squaring the market share of every firm in the industry and summing them:

$HHI = S_1^2 + S_2^2 + S_3^2 \dots$

The mathematical act of squaring the shares serves to highlight and penalize markets dominated by individual giants. Additionally, if the definition of a market boundary is in question, the government employs the SSNIP test (Small but Significant and Non-transitory Increase in Price). This test asks: "If a hypothetical monopolist raised the price by $5\%$, would consumers switch to Product B?" If the answer is affirmative, Product B is considered part of the same relevant market.

Price Discrimination: Transforming Consumer Surplus into Profit

Firms generally find the "single price" model inefficient. If a firm sets a price at $\$50$, it excludes low-income consumers while providing a $\$50$ discount to high-income consumers who were willing to pay $\$100$. Price discrimination is the strategic solution used to capture this "Consumer Surplus." First-Degree Discrimination, also known as Perfect or Personalized Pricing, occurs when a firm charges every individual consumer their exact maximum willingness to pay. Historically, this occurred at car dealerships or bazaars through haggling. Today, companies utilize "Big Data" and algorithms. For instance, an airline might see an IP address searching for flights from a Macbook in a wealthy neighborhood and artificially inflate the price to extract the maximum possible payment. In this scenario, the firm captures $100\%$ of the consumer surplus, and there is zero Deadweight Loss because the firm continues to sell to anyone whose valuation exceeds the Marginal Cost.

Third-Degree Discrimination, or Group Pricing, involves identifying consumers by category (such as Students, Seniors, or Tourists) and charging different prices accordingly. The rule of thumb is to charge a higher price to the group with inelastic demand and a lower price to the group with elastic demand. Consider an exam problem featuring a theme park with two markets. Market 1 (Locals) has a demand of $P_1 = 120 - 2Q_1$, and Market 2 (Tourists) has a demand of $P_2 = 80 - Q_2$, with a constant Marginal Cost of $MC = 20$.

To solve this, first find the Marginal Revenue ($MR$) for each, which is twice as steep as the demand curve: $MR_1 = 120 - 4Q_1$ and $MR_2 = 80 - 2Q_2$. Setting $MR = MC$ for each: for Locals, $120 - 4Q_1 = 20$ results in $Q_1 = 25$ and a price of $P_1 = 120 - 2(25) = \$70$. For Tourists, $80 - 2Q_2 = 20$ results in $Q_2 = 30$ and a price of $P_2 = 80 - 30 = \$50$.

Advanced Pricing Strategies: Self-Selection and Bundling

Second-Degree Discrimination, or Self-Selection, is employed when a firm cannot easily distinguish between rich and poor consumers. They instead offer a "Menu" of options, such as Basic and Premium versions. To prevent wealthy individuals from purchasing the cheaper version, firms use the Incentive Compatibility Constraint. This often involves the intentional sabotage of the lower-tier option. An example is the airline industry, where Economy Class seats are intentionally made uncomfortable to force business travelers to pay for the $\$3,000$ Business Class ticket. The cheap option must be made unpleasant enough that the high-willingness-to-pay consumer feels compelled to upgrade.

Bundling is the practice of selling two or more products together for a single price, such as Microsoft Word and Microsoft Excel. This strategy is most effective when consumers have negatively correlated tastes—for example, when Consumer A values Word highly but not Excel, while Consumer B values Excel but not Word. By selling them as a package, the firm ensures all consumers pay for both products, maximizing total revenue by extracting more value than separate sales would allow.

The Oligopoly Wars: Cournot, Stackelberg, and Bertrand Models

The Cournot Model describes a scenario where firms compete by deciding production quantities simultaneously. The total quantity released into the market determines the price. If both produce too much, the price crashes. To solve a Cournot problem where Market Demand is $P = 100 - Q$ (where $Q = q_1 + q_2$) and $MC = 10$, we start with Firm 1's Total Revenue:

$TR_1 = (100 - q_1 - q_2) \times q_1 = 100q_1 - q_1^2 - q_1q_2$

Taking the derivative yields Marginal Revenue: $MR_1 = 100 - 2q_1 - q_2$. Setting $MR_1 = MC$ gives the Reaction Function: $100 - 2q_1 - q_2 = 10$, or $q_1 = 45 - 0.5q_2$. Since firms are symmetric, $q_1 = 45 - 0.5(45 - 0.5q_1)$, leading to $q_1^* = 30$ and $q_2^* = 30$. Total market quantity is $60$, the price is $P = 100 - 60 = \$40$, and profit for each firm is $(\text{Price} - MC) \times \text{Quantity} = (40 - 10) \times 30 = 900$.

In the Stackelberg Model, competition is sequential rather than simultaneous. Firm 1 acts as the "Leader" and produces first, while Firm 2, the "Follower," reacts. The Leader possesses a First-Mover Advantage by aggressively flooding the market. The Follower, seeing the market is already full, must reduce production to avoid a total price collapse. Mathematically, this is solved using Backward Induction: the Follower's reaction function is substituted into the Leader's demand curve before calculating the Leader's Marginal Revenue.

The Bertrand Paradox describes price-based competition rather than quantity-based. If two gas stations across from each other sell identical fuel and have an identical cost of $\$2$, they will undercut each other's prices (e.g., $\$2.90$, then $\$2.80$) until the price equals the cost ($\$2.00$). The paradox lies in the fact that even with only two firms, the outcome replicates perfect competition and eliminates all economic profit.

Product Differentiation and Hotelling’s Linear City

To escape the "Bertrand Death Spiral," firms use Product Differentiation to ensure consumers do not abandon them for a slightly cheaper rival. This is modeled by Hotelling’s Linear City, where a street of length $1$ has Firm A at $x = 0$ and Firm B at $x = 1$. Consumers living along the street face a transportation cost ($t$) for every step they take. This transportation cost represents brand loyalty or switching costs. For instance, Apple (Firm A) and Samsung (Firm B) represent different "locations" on a preference spectrum. An Apple user faces a high "transportation cost" if they switch to Android (losing iMessage or iCloud).

Because of these switching costs, firms hold local monopoly power over their loyal base. If asked where firms should locate to maximize profits, the Principle of Maximum Differentiation states they should stay as far apart as possible. If they move to the center to steal customers, the distance between them becomes zero, making the products identical and triggering a zero-profit Bertrand price war. Maximizing distance preserves local monopolies and high prices. In equilibrium, price is defined as $P^* = c + t$, showing that brand loyalty ($t$) acts as a profit markup.

Dynamics of Cartels and Collusion

A Cartel is a secret agreement between firms to maintain monopoly-level prices. However, cartels are mathematically unstable due to the Prisoner's Dilemma. If Firm A and B agree on a $\$100$ price, Firm A is tempted to drop its price to $\$90$ to steal the entire market. To maintain a cartel, firms use the Grim Trigger Strategy: they cooperate today, but if one cheats even once, the other will trigger a perpetual, zero-profit price war. A firm will only remain in a cartel if it is sufficiently patient, measured by the discount factor $\delta$. The critical threshold is:

$\delta^* = \frac{\Pi_D - \Pi_M/2}{\Pi_D - \Pi_C}$

(Alternatively written as $\delta^* = \frac{\Pi_{Cheating} - \Pi_{Cartel}}{\Pi_{Cheating} - \Pi_{Cournot}}$). Here, $\Pi_M$ is monopoly profit, $\Pi_D$ (or $\Pi_{Cheating}$) is the one-time gain from cheating, and $\Pi_C$ (or $\Pi_{Cournot}$) is the profit during the punishment phase. If $\delta > \delta^*$, the fear of long-term loss outweighs the short-term greed. Factors that facilitate collusion include high market concentration (fewer firms make coordination easier), high entry barriers (preventing new entrants from undercutting the cartel), transparency (cheaters are caught faster), and multi-market contact (retaliation can happen across multiple business lines).

Mergers and the Problem of Double Marginalization

Horizontal Mergers involve direct competitors (e.g., Ford and Chevy). These involve the Williamson Trade-Off: the downside is increased market power and higher prices (Deadweight Loss), while the upside is increased efficiency through shared technology and lower Marginal Costs. Vertical Mergers involve a company buying its supplier or distributor. Regulators often approve these because they solve Double Marginalization, an inefficiency where both a supplier and a retailer apply their own monopoly markups.

Consider a Leather Supplier with $MC = 20$ and a Shoe Retailer facing demand $P = 100 - Q$. If they are separate, the retailer sets its Marginal Revenue equal to the wholesale price ($w$): $100 - 2Q = w$. The supplier treats this reaction as its own demand curve, with an $MR$ of $100 - 4Q$. Setting $100 - 4Q = 20$ results in $Q = 20$ and a final consumer price of $P = 100 - 20 = \$80$. If they merge, the wholesale price vanishes. The integrated firm sets $\text{standard monopoly MR} = MC$: $100 - 2Q = 20$, resulting in $Q = 40$ and a lower price of $P = 100 - 40 = \$60$. The merger reduces prices, doubles sales, and increases total integrated profit.

Innovation and Arrow’s Replacement Effect

Innovation in IO is often about finding processes that lower Marginal Cost. Kenneth Arrow’s Replacement Effect proves that a firm in a competitive market has a higher incentive to innovate than a monopolist. A monopolist like Blockbuster Video making $\$500$ million only gains an additional $\$100$ million if they innovate to make $\$600$ million; they merely "replaced" their existing profits. A competitive startup like Netflix making $\$0$ profit gains the full $\$600$ million by innovating and becoming the new monopolist. Thus, small firms are the primary drivers of innovation.

Innovation is classified as either Drastic or Non-Drastic. A Drastic Innovation is so superior that the new monopoly price is lower than the old costs of the rivals, driving them bankrupt immediately. A Non-Drastic Innovation is an improvement, but the innovator's ideal monopoly price is still higher than the rivals' old costs. Consequently, the innovator must use "Limit Pricing"—setting the price exactly one penny below the rivals' cost—to maintain the market. Although patents create legal monopolies and Deadweight Loss, they are a "necessary evil." Without the temporary monopoly granted by a patent, firms like Pfizer would never spend billions on R&D, as rivals would simply copy the formula and sell it at cost immediately.