Carlos Upgrade | November 6, 2026 | Sociology and cumulative advantage

Two researchers can submit proposals judged almost equally strong. One receives a grant and the other narrowly misses. Years later, the difference between their careers may be far larger than the original difference between their scores.

That is not necessarily a story about talent. It may be a story about an advantage that creates more advantages.

1. The mechanism

Sociologist Robert K. Merton introduced the Matthew effect in science in 1968. He observed that already-famous scientists could receive disproportionately more credit and attention than lesser-known researchers for comparable contributions. The name refers to a passage in the Gospel of Matthew. Merton was examining recognition, not denying legitimate success.

The wider mechanism is called cumulative advantage. An early edge can become a resource that helps create the next edge:

Recognition → attention and access → more opportunities → further recognition.

This can produce a widening gap even when initial differences are small. But the arrows are not automatic. Resources may be wasted, newcomers can outperform incumbents, and some systems deliberately counterbalance existing advantages.

Thomas DiPrete and Gregory Eirich's 2006 review distinguishes several versions of cumulative advantage. It is no universal law.

2. Outside CPG: a real funding study

In a 2018 study published in PNAS, Thijs Bol, Mathijs de Vaan and Arnout van de Rijt examined a Dutch research funding program. They compared early-career applicants who scored just above and just below a funding cutoff. Their evaluation scores were nearly identical.

Over the following eight years, the narrowly successful group accumulated about €180,000 more in later research funding on average than those who narrowly missed, excluding the original grant. The authors also found that rejected applicants were less likely to apply for subsequent grants. That is an important detail: cumulative advantage can grow not only because gatekeepers favor past winners, but because early setbacks change who keeps competing.

This near-cutoff design was not a randomized trial. One funding system cannot represent every career.

3. A beverage decision

Consider a hypothetical beverage brand in 30 stores, compared with a newer challenger in 10. The established brand has more chances to be discovered, more sales history to show retailers, and potentially a better argument for future space. Its advantage may compound.

But higher total sales do not prove that shoppers prefer its product. Distribution, shelf position, price, promotion and recognition all affect what can be observed.

Before awarding all the next display space to whichever brand sold the most units, I'd compare performance in genuinely similar stores. Measure sales per available selling location, first-time trial, repeat behavior where measurable, stockouts, and the cost of securing those placements. Run a small matched-store test if the decision justifies it. Do not confuse equal opportunity with a promise of equal outcomes.

4. The assumption and the counterargument

Assumption to challenge: The brand getting the most opportunities must be the best brand.

It might be. Retailers often allocate scarce space to products with genuine demand, reliable supply and stronger economics. Existing success can be valuable information, not merely favoritism. Neither dismiss nor idealize market leaders.

The sharper question is whether your evaluation separates performance given an opportunity from the number and quality of opportunities already received.

5. Your mental operator

Use Separate merit from momentum. When one option dominates, ask three things: What did it achieve? What access did it already have? What would we expect under more comparable conditions? Then look for an observation that could disprove your favorite explanation.

6. Reasoning exercise: answer before reading on

Illustrative numbers, not research data: Brand A sells 20 cans per store each week in 30 stores. Brand B sells 40 cans per store each week in 10 stores. Which sells more total cans? Which has higher per-store velocity? Have you proved which is a better product?

Worked answer

A sells 20 × 30 = 600 cans a week. B sells 40 × 10 = 400 cans. A sells more overall, but B sells twice as many per store. That still does not prove B is the better product. Store selection, traffic, shelf position, pricing and inventory may differ. We have identified a comparison worth investigating, not a causal verdict.

Keep this: When success attracts more chances to succeed, ask whether you're measuring quality, access, or the two tangled together.

Research links

• Merton, R. K. (1968), The Matthew Effect in Science: https://doi.org/10.1126/science.159.3810.56
• DiPrete & Eirich (2006), Cumulative Advantage as a Mechanism for Inequality: https://www.annualreviews.org/content/journals/10.1146/annurev.soc.32.061604.123127
• Bol, de Vaan & van de Rijt (2018), The Matthew effect in science funding: https://doi.org/10.1073/pnas.1719557115

Educational editorial content. Hypothetical examples are labeled in the original edition.