A Single Model of the Young-Firm Opportunity

Young-firm employment is now best understood through two linked states: what workers know about the firm’s potential and how easily they can recover if the job goes badly. This chapter connects those states into one model of the opportunity, then turns the model into predictions that can be checked in data. The practical judgment is that firm age alone should predict little; the interaction between short evidence, observed performance, and outside options should predict much more.

Two states decide what a young-firm job is worth

The first state is informational. A young firm has produced too few observations for workers to know whether its early performance reflects a persistent capability or temporary circumstances. A high result can be genuinely encouraging without being conclusive. A low result can be genuinely concerning without proving that the business has no future. Firm age tells us how much history exists; past-average performance tells us which direction the history points; signal quality tells us how much confidence the history deserves.

The second state is positional. A worker’s cost of being wrong depends on what happens after the match fails. A worker with accessible alternatives can search while employed or find another job after a separation. A worker facing scarce vacancies, a slow search, or a large loss of income has a weaker outside option. The same uncertain young firm is therefore a more expensive choice for the second worker, even if both workers receive the same offer.

These states interact rather than operate in parallel. Information determines the expected future value of the firm. Mobility determines how heavily the worker discounts that value if it turns out to be wrong. The firm then competes through current wages and continuation promises, while workers decide whether to join, remain, or search elsewhere. Hiring and retention are the bridge from an individual belief to the firm’s growth path.

This gives the opportunity two distinct axes. One asks, “How convincing is the firm’s evidence?” The other asks, “How recoverable is the worker’s mistake?” A young firm can have strong evidence but a difficult labor market around it, producing a need for a larger premium. It can have weak evidence but a fluid labor market, so workers can test the match without bearing the full cost of failure. Treating these cases as one category called “start-up risk” hides the mechanism that determines the actual employment terms.

flowchart TD A[Young firm: short, noisy history] --> B[Belief about underlying potential] A --> C[Worker outside option and mobility] B --> D[Expected future job value] C --> D D --> E[Wage and continuation terms] E --> F[Join, stay, or search] F --> G[Hiring, retention, or exit] G --> H[Growth and aggregate productivity]

This chain prevents several category errors. A wage is not the same as a belief, because the firm can respond to a belief through other contract terms. A belief is not the same as a realized outcome, because the worker may be wrong about the firm’s future. A firm’s growth is not the same as a worker’s welfare, because growth can coexist with layoffs, unequal pay, or a difficult outside market. The model is useful precisely because it keeps these links visible without treating them as one outcome.

It also explains why the relevant comparator is an apparently similar mature firm, not an abstract “safe job.” The comparison holds current performance and size as close as possible, then asks what remains because the histories differ. That remainder is the informational age effect.