📈 economics

Fertility Choices with Career Risk

This paper demonstrates that limited-commitment frictions in household bargaining, where career risks from parenthood are realized after fertility decisions, disproportionately expose wives to earnings risk and reduce desired family size, but targeted insurance against the uncertainty of these penalties can effectively restore higher fertility levels without subsidizing the expected cost of children.

Original authors: Ruiwu Liu

Published 2026-08-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ruiwu Liu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Most people assume that the decision to have children is a calculation of costs and benefits made with a clear picture of the future. In economics, this is often modeled as a couple weighing the joy of a new life against the known price of diapers, time, and lost wages. But this view misses a crucial element of modern life: the future is rarely clear. When a couple decides to start a family, they do not know exactly how much their careers will suffer. They do not know if the mother will face a steep drop in pay, a stalled promotion, or a permanent shift in her professional path. This uncertainty is not just a background worry; it is a central force that shapes how many children a family actually wants.

This paper explores that specific uncertainty. It looks at a situation where a couple agrees on a family size before they know the true cost of parenthood for the mother's career. Once the children are born and the career loss becomes real, the couple must renegotiate how they share their resources. Because they cannot make a binding promise to fully compensate the mother for whatever damage her career takes, the risk of that damage falls disproportionately on her. The researchers built a model to see how this lack of a perfect safety net changes the number of children a couple chooses to have, and whether insurance against this specific risk could change their minds.

The study finds that the fear of an unpredictable career hit is enough to shrink family sizes, even if the average cost of having children stays the same. In the model, when the risk of a severe career penalty increases, couples choose to have fewer children. This happens not because they expect to be poorer on average, but because they are afraid of the worst-case scenario. The model shows that this fear is amplified by the way couples share money after the children arrive. If the mother's future income determines her ability to leave the marriage or live on her own, she bears a larger share of the career risk. When the risk is high and she is the one most exposed to it, the couple agrees to have fewer children to avoid the potential disaster.

The researchers tested this idea with a numerical simulation. They set up a scenario where a couple might want three children if the future were certain. When they introduced a realistic amount of career risk, the couple's desired family size dropped to two. If the risk increased further, the number fell to just one. This drop happened even though the average cost of raising a child did not change; only the uncertainty around that cost increased. The study suggests that the problem is not the cost itself, but the inability to insure against the unpredictable part of the career loss.

The paper then compares two different ways a government might try to help. One is a standard subsidy, like a cash payment for every child, which lowers the average cost of raising a family. The other is a form of insurance that protects the mother against the unexpected part of her career loss, without changing the average cost. The simulation shows that while both methods can encourage larger families, they work in different ways. A cash subsidy helps by making children cheaper on average. The insurance, however, helps by calming the fear of a career catastrophe. In the specific scenario tested, a policy that insured half of the unexpected career loss was enough to restore the couple's desire for a three-child family, whereas a cash subsidy of a similar size did not cross the threshold to bring the family size back up.

This distinction is important because it suggests that policies focused only on the average cost of children might miss a key lever. If the main barrier to having more children is the fear of an unpredictable career hit, then a policy that smooths out that risk could be more effective than a simple cash handout. The model does not claim this is a universal rule for every family, but it does show that when the risk of a career penalty is high and concentrated on the mother, insuring that risk can change the family planning decision in a way that ordinary subsidies cannot.

The author is careful to note that their work is a theoretical model supported by numerical examples, not a final answer based on real-world data. They have not yet tested these ideas against actual birth records or specific government programs. Their goal was to show that the mechanism exists: that limited ability to commit to future compensation creates a wedge that reduces fertility, and that insurance against the distribution of career penalties is a distinct policy tool. The next step, they suggest, is to take this framework and apply it to real data, measuring exactly how much career uncertainty varies across different jobs and how much it actually affects family size in the real world. Until then, the study stands as a clear demonstration that the fear of the unknown can be just as powerful as the known cost in the decision to bring a child into the world.

Technical Summary: Fertility under Career Risk: Limited Commitment and Insurance against the Child Penalty

Problem Statement
Existing household fertility models typically summarize the "child penalty" by its expected level, assuming fertility decisions and transfers occur under complete information. This paper addresses a distinct margin: couples often decide on fertility before the persistent career loss associated with parenthood (specifically for the mother) is realized. Once the loss occurs, household resources must be renegotiated under limited commitment, as ex ante agreements regarding the division of future losses are not fully enforceable. The central problem is how this uncertainty, combined with the inability to fully commit to state-contingent transfers, distorts fertility decisions and the incidence of career risk within the household.

Methodology
The paper constructs a generalized-Nash fertility bargaining model with the following structural features:

Timing: Spouses bargain over fertility ( $n$ ) before the realization of a persistent career loss per child, $X = \mu + \varepsilon$ . After $X$ is observed, household consumption is renegotiated.
Limited Commitment: Post-birth consumption allocations depend on outside options, which are functions of individual market earnings. The wife's post-birth income is $y_w(n, X) = a_w - Xn$ .
Incidence Mechanism: A parameter $\lambda \in [0, 1]$ measures the extent to which own earnings protect an individual's outside option during renegotiation. This determines the "incidence" of the career loss: a share $\theta_w$ falls on the wife and $\theta_m$ on the husband. Crucially, limited commitment shifts a portion of the risk from the husband to the wife as $\lambda$ increases.
Preferences: Spouses derive utility from children ( $b_i \log(1+n)$ ) but bear costs including direct resource costs, non-market burdens, and the expected career loss. Preferences over the random payoff are modeled using a mean-variance representation with constant absolute risk aversion ( $\eta_i$ ), treating the certainty equivalent as the cardinal surplus for bargaining.
Solution Concept: The equilibrium fertility level $n^*$ maximizes the Nash product of the spouses' ex ante gains, subject to individual participation constraints ( $G_i(n) \geq 0$ ).

Key Contributions and Mechanisms
The paper identifies a "risk-sharing distortion" arising from limited commitment that is absent in deterministic participation-constraint models or full-commitment benchmarks.

Risk Concentration: Limited commitment does not create the aggregate career loss but determines who bears it. Because outside options depend on realized earnings, the spouse with higher earnings sensitivity (typically the wife) bears a disproportionate share of the risk, even if the aggregate expected resources are unchanged.
Incidence Result: The model derives closed-form expressions for the loss incidence shares ( $\theta_w, \theta_m$ ). It shows that as the loading of own earnings into outside options ( $\lambda$ ) increases, the wife bears a larger share of both the expected loss and the risk.
Policy Distinction: The paper distinguishes between two policy instruments:
1. Insurance ( $q$ ): A state-contingent contract insuring the unexpected component ( $\varepsilon$ ) of the career loss. This reduces the variance of the penalty without changing its mean.
2. Subsidy ( $s$ ): A conventional per-child transfer that reduces the expected cost ( $\mu$ ) but leaves the risk variance unchanged.

Key Results

Existence and Uniqueness: A unique positive fertility agreement exists if and only if both spouses have positive marginal gains at zero fertility. The solution is characterized by the first-order condition of the Nash product.
Comparative Statics on Risk:
- An increase in the expected career loss ( $\mu$ ) reduces agreed fertility (expected resource effect).
- A mean-preserving increase in career-loss risk ( $\sigma^2$ ) also reduces agreed fertility. This "risk effect" persists even when expected household income is held constant.
The Role of Limited Commitment ( $\lambda$ ): The effect of increasing the sensitivity of outside options to own earnings is ambiguous in sign generally but negative under specific conditions. Fertility falls if the wife's participation margin is more sensitive to the shift in risk than the husband's offsetting gain. This occurs when the wife is more risk-averse, bears a larger non-market burden, or is closer to her participation boundary.
Insurance vs. Subsidy:
- Insurance against the unexpected component of the career loss raises fertility by compressing the distribution of the penalty.
- A child subsidy raises fertility by lowering the expected resource cost.
- The two instruments are not equivalent unless risk is absent or households are risk-neutral. Insurance can restore fertility levels even when the mean penalty remains high.
Full-Commitment Benchmark: In a benchmark where spouses can commit to state-contingent transfers ex ante, risk is shared efficiently based on relative risk aversion. The decentralized limited-commitment equilibrium concentrates risk on the more risk-averse spouse (the wife), leading to inefficiently low fertility compared to the full-commitment optimum.

Numerical Illustrations
The paper provides a discrete-parity exercise (integer family sizes) to demonstrate the magnitude of these effects:

Risk Impact: Rising career risk (increasing $\sigma$ ) shifts the desired family size from three children to two, and then to one.
Policy Impact: Partial earnings-loss insurance can restore the three-child outcome at risk levels that would otherwise reduce fertility to one. In contrast, a conventional child subsidy requires a larger magnitude to achieve the same parity shift in the illustrative example.
Interaction: The paper notes numerically that the proportional effect of insurance may be larger where post-birth resources track own earnings more closely (higher $\lambda$ ).

Significance and Claims
The paper claims to identify a specific policy margin unavailable in deterministic models: insurance against the distribution of the child penalty.

It argues that the "uninsurable career-risk wedge" is a distinct mechanism from standard bargaining power models.
The contribution is described as "narrower and analytically transparent": isolating the risk-sharing distortion, deriving its incidence, and comparing insurance against risk with ordinary subsidies.
The author explicitly states that the current numerical exercises are illustrative, not calibrated. The paper positions itself as completing the "analytical core" and verifying mechanisms, with the next required steps being empirical discipline (estimating the shock distribution) and a quantitatively specified policy environment for journal submission.
The primary headline is not that bargaining power affects fertility (a crowded field), but that couples choose fertility before the long-run career cost is known, limited commitment concentrates this risk on the mother, and insurance against the career-loss distribution cannot generally be replicated by a baby bonus.

Technical Summary: Fertility under Career Risk: Limited Commitment and Insurance against the Child Penalty

More like this