Gibrat's Law: Business and Network Growth Analysis

Gibrat's Law: The Law of Proportionate Effect

At its core, Gibrat's Law is elegantly simple. It states that the proportional growth rate of an entity is independent of its absolute size.

Imagine two firms, one with $100,000 in revenue and another with $100 million. Gibrat's Law posits that both have the same probability of growing by 10% in the next year. The growth is proportionate. Mathematically, if $S_t$ represents the size of the entity at time $t$, its size at the next step is:

where ${ϵ}_t$ is a random variable representing the growth rate, whose distribution is independent of $S_t$. Over many time steps, this multiplicative process tends to produce a log-normal distribution of sizes across a population of such entities. As a statistician, you'll recognize the log-normal distribution as the result of many independent, random multiplicative factors.

The real power of Gibrat's Law lies not in its universal truth—it is frequently violated—but in its utility as a baseline. It forces us to ask a critical question: is growth in this system truly random and size-independent, or are there other forces at play?

Application 1: Growth of a Lingerie Business

Let's consider your experience as a lingerie designer. Imagine you've started a boutique lingerie brand. Your brand is small and agile. In the same market, there is a large, established lingerie conglomerate.

Scenario: According to a strict interpretation of Gibrat's Law, your boutique brand and the large conglomerate have the same expected proportional growth rate. If the average growth rate in the industry is 5%, both your company and the giant are equally likely to hit that mark.

Evaluation: How does this hold up in the real world? Is a small startup's growth dynamic truly the same as a market leader's? Let's look at some evidence.

The study's findings directly contradict Gibrat's Law. This is a common result in business economics. Why might this be?

• Law of Large Numbers: It is arithmetically easier for a business with $50,000 in annual sales to double to $100,000 than it is for a $50 million company to double to $100 million. The smaller firm is starting from a lower base.
• Market Saturation: The large conglomerate may already serve a majority of the market, making significant further expansion difficult. Your boutique, however, can achieve high percentage growth by capturing even a tiny sliver of the market.
• Agility: Your smaller brand can pivot quickly to new trends, materials, or marketing strategies. The large firm may be bogged down by bureaucracy and established processes.

In this context, Gibrat's Law is a flawed predictor. For your lingerie business, you would likely reject the null hypothesis that your growth potential is the same as the market leader's. In fact, you might build your strategy around the expectation of faster proportional growth in the early stages, a common feature of startups that successfully find a market niche.

Application 2: Expansion of a Polyamorous Network

Now let's apply this to our second scenario: the "Vee" relationship, where a husband supports his wife’s relationship with her lover. This vee is part of a small, tight-knit social circle of a dozen people who are aware and supportive of their non-monogamous lifestyle. In the same city, there is a large, established polyamorous community with hundreds of members, regular meetups, and a strong online presence.

Scenario: Gibrat's Law would predict that the small friend circle and the large community have the same proportional chance to grow. If an average social group in this space grows by 20% a year, Gibrat's law would suggest both are equally likely to do so, meaning the small group adds 2-3 people and the large group adds dozens.

Evaluation: Is the growth of social networks as simple as proportionate, random additions? Or does size itself influence the ability to attract new members? A research paper on the growth of online social groups provides a sophisticated answer.

The paper demonstrates that for social groups, size matters, but often in the opposite way to our business example. Large social groups often grow proportionally faster than small ones. This phenomenon is better explained by models of preferential attachment, often summarized as "the rich get richer."

In network science, the probability of a new node connecting to an existing node depends on that node's current number of connections (its degree, $k$). This relationship is often modeled as a power law: $\pi (k)\sim k^{\alpha }$. Gibrat's law is conceptually similar to the case where $\alpha =0$ (attachment is random), but social networks rarely behave this way.

The image below illustrates what happens when $\alpha$ is not zero.

Let's analyze the growth of the polyamorous community through this lens:

• Linear Growth ($\alpha =1$): "Rich get richer." This is the classic model of preferential attachment. A large community with 500 members is far more visible and has more connections than a small circle of 12. It will naturally attract a disproportionately larger share of newcomers seeking community. This leads to a scale-free network structure, with a few very large hubs (the big communities) and many small, peripheral groups (like the vee's immediate circle).
• Superlinear Growth ($\alpha >1$): "Winner-take-all." If the large community is the only major hub in the city, its advantage might be even more pronounced. It could become a Schelling point, attracting nearly all newcomers and leading to a stark hub-and-spoke topology where smaller, independent groups struggle to form or grow.
• Sublinear Growth ($0<\alpha <1$): Diminishing returns to size. While larger groups are still more attractive, the effect is muted. Perhaps the large community becomes cliquey, impersonal, or hard to navigate for newcomers. In this case, a medium-sized group might be seen as the "sweet spot," or a small, intimate circle might be more appealing to some, giving it a chance to grow.

For the polyamorous vee, evaluating their network's growth potential means rejecting Gibrat's Law and instead asking: what is the value of $\alpha$? If $\alpha \ge 1$, the large community will likely dominate growth. If $\alpha <1$, their smaller, more intimate circle may have a better chance of attracting members who seek that specific dynamic.

Conclusion

Gibrat's Law provides a crucial, if often inaccurate, starting point for analyzing growth. It proposes a world where size doesn't confer an advantage or disadvantage. By testing this simple model against reality, we uncover the more complex and interesting dynamics that truly govern a system.

Key Takeaways:

• Gibrat's Law (Law of Proportionate Effect): A baseline model asserting that an entity's proportional growth rate is independent of its size. It often leads to a log-normal distribution of sizes.
• Application to Business: The law is frequently violated, as smaller, more agile firms often exhibit faster proportional growth than large, saturated incumbents.
• Application to Social Networks: The law is also violated here, but often in the opposite direction. Due to preferential attachment ("rich get richer"), larger social groups tend to attract a disproportionate share of new members and grow proportionally faster.
• Strategic Value: The primary use of Gibrat's Law is as a null hypothesis. Its failure forces you to investigate the true relationship between size and growth in your specific domain, be it commercial or social.

In our next lesson, we will explore Jevons' Paradox. Having examined quality and growth, we will turn to a deeply counterintuitive principle of efficiency, which suggests that improvements in resource efficiency can paradoxically lead to increased consumption of that resource.