Introduction: This Is Not an Article About Families. It Is an Article About Stepfamilies.
Begin with the boundary, because the boundary is where most family policy arguments go wrong. This essay is not about families in general. It is about the specific, second-best, and unavoidable configuration that every low-fertility society now has to manage: the stepfamily.
The priority ordering must be stated before anything else, because a policy that will not name its first preference is a policy that has already conceded the argument. The primary object of state family subsidy should be the freshly married couple raising their own shared biological children. That household is the cleanest case: it produces the next generation of taxpayers, it carries the lowest structural risk of the configurations available, and it is the only form in which the common-child stabilizer is present from the first day rather than reconstructed later. Everything that follows is subordinate to that claim, not a substitute for it.
But stepfamilies are a fact, and facts are not policy preferences. Roughly a quarter to a third of children in the EU will spend part of their childhood in a household that includes a step-parent, and the number is rising with every cohort of separations. No government has ever reversed that trend by disapproving of it, and none will. The question is therefore not whether the state will deal with stepfamilies. It will. The question is whether it deals with them in their most stable and safest forms—formalized, shared-child-bearing, legally recognized, and supported—or whether it continues to penalize those forms while subsidizing the fragile ones. The argument of this essay is that it currently does the latter, and that the fix is arithmetic rather than sentimental.
The Fiscal Externality Nobody Wants to Name
Children are a public good with private costs. This is not a sentiment. It is an accounting fact, and it is the foundation of any honest family policy.
A child born in the EU today will spend forty years paying into a pay-as-you-go pension system that funds the retirement of people they will never meet. They will staff the hospitals, the care homes, the tax offices, and the army. They will generate the tax base on which every public service depends. And the cost of producing them—the housing, the food, the childcare, the eighteen years of uncompensated labor—is borne almost entirely by the parents.
This is a textbook externality. The social return on a well-raised child exceeds the private return to the parents, because the state captures a large share of the upside through taxation and the pension system while bearing almost none of the cost. Economists have a word for goods whose production benefits people who did not pay for them: underprovided. That is what children are in every low-fertility country in Europe. They are underprovided because the people who bear the cost do not capture the benefit, and the people who capture the benefit do not bear the cost.
The state therefore has a legitimate fiscal interest in the question of which family forms produce children well, and which ones do not. This is not moralism. It is the same logic that justifies subsidizing education, vaccination, and infrastructure. The question is not whether the state should have a family policy. It already has one—an enormous, expensive, incoherent one, distributed through the tax code and the benefit system. The question is whether the existing policy is pointed in the right direction.
The argument here is that it is not. The state currently penalizes exactly the family form it says it wants, subsidizes the one it says it wants to discourage, refuses to condition any of it on the commitments that make the difference, and ignores the single design variable that separates the stable stepfamily from the fragile one. The fix, while contested, is more tractable than the political class pretends.
Part One: The Single Mother as Fiscal Asset and Fiscal Liability
Start with the accounting, because the moral case follows from it.
A single mother raising a child is producing the next generation of taxpayers. She is doing the most fiscally valuable work a citizen can do, and she is doing it alone. The AROPE rate for children in single-parent households stood at 42.2% in 2024, against roughly 11% for two-parent families. Single-parent households face a poverty risk of over 32%, more than double the overall population. These are not abstractions. They are the numbers that determine whether the child born to a single mother in 2026 becomes a net contributor or a net recipient over the course of their life.
The state already subsidizes her, heavily and unconditionally. Child benefit, housing allowance, child maintenance guarantees, tax credits, social assistance, subsidized childcare. In Finland alone, a single mother with one child receives child benefit (€94.88), the single-parent supplement (€73.30), and the child maintenance allowance (€197.71)—roughly €366 per month before any means-tested support. Germany pays €259 per month per child plus a means-tested supplement of up to €297 for single parents earning at least €600. France pays €96.27 per child plus €200.78 per child through the allocation de soutien familial. Ireland pays up to €254 per week through the One-Parent Family Payment.
The subsidies exist for good reasons. They reduce child poverty, they stabilize households, and they prevent the intergenerational transmission of disadvantage. Nobody is arguing for their abolition.
But they are unconditional on the variable that matters most: the presence of a second committed adult in the household. And that is the design flaw.
Part Two: The Marriage Premium—What It Is and What It Is Not
The empirical literature on family structure and child outcomes is one of the most contested in the social sciences, and the contest is instructive.
The raw correlation is large. Children raised in married-parent families have better outcomes on almost every measure: higher educational attainment, lower rates of delinquency, lower rates of early sexual activity, better mental health, higher household income. McLanahan and Sandefur’s Growing Up with a Single Parent (1994) remains the canonical statement of the gap, and McLanahan, Tach, and Schneider’s 2013 review in the Annual Review of Sociology concluded that the evidence for a causal effect of father absence is real, though smaller than the raw correlation suggests.
The causal estimate is smaller and contested. The problem is selection. People who marry are different from people who do not—they are more educated, more religious, more prosocial, more likely to come from stable families themselves. Much of the raw marriage premium is the premium on the kind of person who marries, not on marriage itself. When researchers use methods that account for selection—sibling comparisons, instrumental variables, natural experiments—the effect shrinks substantially.
The male marriage wage premium is real and partially causal. Korenman and Neumark’s 1991 study in the Journal of Human Resources found that married men earn more than unmarried men, and that a substantial portion of the gap survives controls for selection. Antonovics and Town’s 2004 work in the American Economic Review decomposed the premium and found that it is partly a productivity effect and partly a selection effect. The mechanism is not mysterious: marriage is associated with specialization, reduced labor market mobility, and—crucially—the expectation of long-term provisioning, which changes the man’s own behavior.
This is the part of the argument that survives scrutiny. Commitment changes behavior. A man who has made a legally binding, publicly witnessed, socially enforced commitment to a woman and her children behaves differently from a man who has not—not because the piece of paper is magic, but because the piece of paper changes the incentive structure, the expectations of both parties, and the cost of exit.
That is the mechanism. The state should care about it because it is the mechanism through which the next generation of taxpayers is produced.
Part Three: The Stepfather—The Underrated Variable, and the Commitment Filter
Now the uncomfortable part, and the part the essay’s thesis depends on.
A man who marries a single mother is doing something the state should value enormously. He is not merely adding an income. He is adding a second committed adult to a household that otherwise has one. He is providing the paternal investment that the biological father is not providing, and he is providing it under conditions that make the provision credible—legal marriage, shared residence, public commitment, and financial entanglement.
The evolutionary anthropology here is honest and it cuts both ways. Anderson, Kaplan, and Lancaster’s work on paternal investment found that stepfathers invest less than genetic fathers on average, and Daly and Wilson’s work on the “Cinderella effect” documented elevated risk of abuse and homicide in stepfamilies relative to biologically intact families. These are real findings, and they must be stated plainly. Stepfathers are not a uniform category of saint. The distribution of stepfather investment has a lower floor than the distribution of biological father investment, and the state should not pretend otherwise.
But four qualifications matter, and they are routinely ignored.
First, the absolute risk is small but significant, and the smallness is not a comfort. The Cinderella effect is a relative risk, not an absolute one. The overwhelming majority of stepfathers do not abuse their stepchildren. But the elevated rate is real, it is not fully explained by selection, and the harm when it occurs is catastrophic and irreversible. A population-level elevation in a rare outcome is still a population-level elevation in a rare outcome; the fact that most stepfathers are not abusers tells the child who is being abused nothing useful. Some of the effect is selection—the characteristics of the men who enter stepfamilies and of the households that form them—but a residual remains after every adjustment the literature has attempted, and the policy design must treat the risk as a cost to be managed and priced, not a statistic to be waved away.
Second, the mechanism that reduces stepfamily instability is the “common child.” The most robust finding in the stepfamily literature is that stepfamilies that produce a shared biological child are substantially more stable than those that do not. The shared child converts the stepfather from an investor in someone else’s genetic line into an investor in his own—and it gives the couple a shared project that outlasts the strains of stepfamily formation. This is the single most important design variable in the whole essay, and it is the one the state currently ignores entirely.
Third, the alternative is not the biological father. The comparison is not stepfather versus intact biological family. The comparison is stepfather versus no second adult in the household at all. And on almost every measure that matters to the state—household income, housing stability, maternal stress, educational supervision, delinquency risk—the married stepfather household outperforms the single-mother household. The Cinderella effect is a real risk to be managed. It is not a reason to prefer the absence of a committed adult as the default.
Fourth, and this is the one the objection always misses: the commitment process itself is a filter. The elevated risk documented in the stepfamily literature is not evenly distributed across all stepfathers. It is concentrated in the transient, unformalized, low-commitment end of the distribution—the boyfriend who moves in and out, the serial cohabitation arrangement, the man who has no legal or financial stake in the household and no public accountability for what happens inside it. The literature on family instability and child maltreatment consistently finds that the number of maternal partner transitions is a stronger predictor of risk than the presence of a stepfather as such. It is the churn that harms, not the role.
The formal commitment process screens against exactly the disposition that produces that churn. A man who is willing to marry—to sign a legal contract, to make a public vow, to accept financial entanglement, to submit to the possibility of divorce proceedings and property division, to be named in school records and medical forms—is a man who has demonstrated a willingness to bear costs for the household. That willingness is precisely the opposite of the exploitative disposition. The abuser selects for arrangements with low exit costs and low accountability. Marriage raises both. It does not eliminate the risk—no institution does—but it pre-selects against the type, and it makes the type’s behavior visible and actionable in ways that unformalized cohabitation does not.
This is a selection argument, and it must be labeled as one. It is not proof that married stepfathers are safe. It is the claim that the commitment process is a screening device, and that the state, by refusing to recognize or reward it, is failing to use the cheapest screening device it has. The woman choosing between a man who will marry her and a man who will not is already running this filter instinctively. The state’s benefit system currently runs it in reverse—taxing the marriage and subsidizing the cohabitation—and then wonders why the fragile arrangements proliferate.
The state’s interest in stepfathers, then, is real but conditional. It should encourage the formation of committed stepfamilies, recognize them legally, and support the conditions—shared children, stable housing, counseling, legal clarity—under which they are most likely to last. The commitment is not decoration. It is the filter.
The Reproductive Window Is a Design Variable
The commitment-filter argument has a corollary that the policy design must absorb: the fiscal rationale for subsidizing a stepfamily is not uniform across the reproductive life course. The state is buying two different things in two different configurations, and it should pay for them differently.
Where the husband is in his reproductive window, the state is buying a producer—a household with the capacity to generate a shared child, which is the stabilizer the stepfamily literature identifies as the single most important determinant of durability, and which is the source of the next generation of taxpayers the state actually cares about. This is the configuration where the subsidy has the highest expected fiscal return, because it is the only configuration in which the household can convert the commitment into a new net contributor.
Where the husband is clearly out of his reproductive window, the state is buying something different: a stabilizer. He cannot produce the shared child. The common-child mechanism is structurally unavailable to him. Which means the marriage is doing all of the stabilization work on its own—and that makes him, counterintuitively, the purest case of the commitment filter in the entire taxonomy.
Consider what the old stepfather cannot do and what he therefore is. He cannot buy the stabilizer; he can only be the stabilizer. He is not going to produce a child he will not raise to adulthood—the fiscal downside risk that hangs over the young stepfather is simply absent. And the churn risk that concentrates in young, transient, unformalized arrangements—the serial boyfriend, the rotating cohabitant, the man with no legal or financial stake—is at its minimum, because the population of men willing to formally marry at 60 and take on another man’s teenage children is not the population of men who cycle through households. The commitment filter is doing 100% of the work, with none of the offsetting risk.
The current system treats both configurations identically: it taxes the marriage and subsidizes the separation, regardless of whether the marriage is capable of producing the next taxpayer or is merely providing a stable second adult for the children who already exist. That is the design flaw the tiering below corrects.
Part Four: Cohabitation Is Not Marriage—The Stability Data
Here the evidence is strongest, and the argument for the state’s refusal to sponsor cohabitation is actuarial rather than moral.
Cohabiting parents are substantially less stable than married parents. Brown’s 2004 work in the Journal of Marriage and Family found that children in cohabiting-parent families experience more family instability and worse outcomes than children in married-parent families. Manning and Brown’s 2015 study found that children in cohabiting households have substantially lower economic well-being than children in married households—even after controlling for income, education, and other characteristics. The dissolution rate for cohabiting parents is roughly two to three times the dissolution rate for married parents, depending on the cohort and the country.
Cohabitation is heterogeneous, and the heterogeneity matters. The literature distinguishes “cohabitation as a stage” (a prelude to marriage, associated with younger and more educated couples) from “cohabitation as an alternative” (a substitute for marriage, associated with lower education, lower income, and higher instability). The first is relatively stable. The second is not. And the second is where most of the children in cohabiting families now live.
The causal question is genuinely contested. How much of the instability gap is caused by the absence of legal marriage and how much is caused by the kind of people who cohabit rather than marry? The honest answer is that both are operating, and the literature has not cleanly separated them. Some of the stability advantage of marriage is selection. Some of it is the commitment mechanism—the legal, financial, and social costs of exit that make people work harder before they leave. The best estimates suggest that a meaningful fraction is causal, particularly for couples with children, but the precise fraction is not established.
The Nordic evidence complicates the picture in an instructive way. In Sweden and Norway, where cohabitation is culturally and legally normalized, cohabiting parents are much closer to married parents in stability and in child outcomes than they are in the United States or the UK. This is important because it suggests that the stability gap is not entirely intrinsic to the absence of a marriage certificate. It is partly a function of the legal and social recognition the partnership receives. In the Nordics, cohabitation has many of the legal attributes of marriage—inheritance, custody, parental responsibility, tax treatment—and the gap narrows accordingly.
This cuts against a crude “marriage only” position. The lesson is not that cohabitation is as good as marriage. It is that the legal recognition of commitment is what matters, and marriage is the institution through which that recognition has historically been delivered. A jurisdiction that extends legal recognition to cohabitation has, in effect, created a second marriage—and should not be surprised when it behaves like one.
The defensible position, then, is not that cohabitation is inherently inferior. It is that the state should not extend the full package of family subsidies to a relationship form that is, on average, less stable and less committed—and that its current policy does precisely the opposite.
Part Five: The Couple Penalty—How the State Punishes Exactly What It Says It Wants
This is the strongest part of the case, and it is the part that is least contested.
The tax and benefit system currently penalizes partnership. This is not a conspiracy. It is the arithmetic consequence of means-testing benefits on household income. When a single mother on housing benefit, child tax credit, and income support marries or partners with a man who earns a modest wage, the household income rises—and the benefits are withdrawn. The marginal effective tax rate on the man’s earnings can exceed 70% or even 80% in some configurations. The couple is worse off financially than the two individuals were separately.
The Institute for Fiscal Studies documented this in the UK as the “couple penalty.” Adam and Brewer’s 2004 analysis showed that the UK tax and benefit system imposed a significant penalty on forming a couple, and subsequent reforms have not eliminated it. The same structure exists across the EU, in different forms and at different magnitudes. In Finland, the general housing allowance is calculated on household income, so a partner’s earnings reduce the allowance. In Germany, the Kinderzuschlag has a household income test. In France, the allocation de soutien familial is withdrawn when a partner enters the household.
The state is doing the exact opposite of what the fiscal externality argument recommends. It is taxing commitment and subsidizing separation. It is making it financially irrational for a single mother to marry a man who would provide the second committed adult in the household—the very outcome the state says it wants.
This is not a small distortion. It is the primary mechanism through which the state shapes family formation, and it is pointed in the wrong direction. The first-order policy recommendation is not to add a new subsidy. It is to stop taxing the one the state already wants.
Part Six: The Evidence on Marriage Subsidies—What Works and What Doesn’t
Here the argument must be honest about a hard fact: the evidence that marriage promotion programs work is weak.
The two largest evaluations are the Supporting Healthy Marriage project and the Building Strong Families project, both funded by the US federal government and evaluated by MDRC.
Supporting Healthy Marriage (Hsueh et al., 2012) tested relationship education for married couples with children. It found modest positive effects on relationship quality and on some measures of couple communication. It found no effect on marriage rates, no effect on divorce rates, and no consistent effect on child outcomes.
Building Strong Families (Wood et al., 2012) tested relationship education for unmarried couples who were expecting or had just had a child—the population closest to the essay’s target. The overall results were null. Worse, for one subgroup—couples with the lowest levels of initial commitment—the program appeared to increase the rate of relationship dissolution, and some child outcomes were worse.
The lesson is important and it is not that marriage is bad. The lesson is that teaching people to communicate does not change whether they marry. Relationship education addresses the soft skills of partnership. It does not address the structural conditions—housing, income, employment, the benefit cliff—that determine whether marriage is a rational choice. The BSF finding that the least-committed couples did worse is a warning: pushing fragile couples into commitment can backfire, and the children pay the price.
What the evidence does support is the structural channel. Removing the couple penalty changes behavior. When the UK introduced the Working Tax Credit and made work pay better for couples, partnership formation among low-income households changed. When Sweden extended parental leave and childcare to cohabiting parents, the gap between cohabiting and married parents narrowed—not because the couples changed, but because the institutional treatment changed.
The policy implication is clear. Do not fund relationship education as a marriage strategy. Fund the removal of the fiscal penalty, and let the commitment follow the incentive. The state cannot make people love each other. It can stop making it expensive for them to stay together. And it should go one step further: commitment should be financially rewarded with direct subsidies, not merely unpenalized. Removing a penalty restores neutrality; a direct subsidy tilts the calculus in favor of the commitment the state says it wants. The two instruments do different work. The penalty removal is the precondition—it stops the benefit system from actively pushing couples apart. The direct subsidy is the positive incentive—it makes the committed household measurably better off than the uncommitted one, so that the couple choosing between cohabitation and marriage is choosing with the state’s weight on the scale rather than against it. The logic is the same one that governs every other subsidized public good: if the behavior generates an externality the market does not price, pay for it. Commitment generates the stability that produces and protects the next generation of taxpayers. The state should therefore pay for commitment directly—through the tiered couple premium described in Part Seven—and it should do so in a form that is visible, predictable, and large enough to matter at the margin. More couples will commit when commitment is rewarded. That is not a hope. It is the straightforward prediction of the incentive structure the policy creates.
Part Seven: The Design—A Tiered Subsidy Keyed to the Reproductive Window
The design rests on two instruments, and it is essential to keep them distinct.
Instrument one: the child benefit. Paid to the household that contains the child, regardless of the parents’ marital status. It exists to reduce child poverty and stabilize the child’s material environment. It is not a marriage subsidy and should not be treated as one. Every child, in every household configuration, receives it.
Instrument two: the couple premium. Paid only to households where the commitment has been legally formalized—and tiered according to the stability profile of the configuration. This is the marriage subsidy, and it is the instrument the state uses to purchase the second committed adult. It is not paid to childless marriages, because there is no externality to purchase. It is not paid to unformalized cohabitation, because there is no commitment to reward. It is paid to married households that include a child or stepchild, at rates that reflect how durable the configuration is likely to be.
Tier 0 — Unformalized cohabitation: child benefit only, no couple premium.
The actuarial case for treating unformalized cohabitation as a lower-commitment form stands. The state should offer cohabitants the legal instruments of commitment—registered partnership, parental responsibility agreements, inheritance rights, next-of-kin status—and then treat the exercise of those instruments differently from their absence. The child benefit flows to the child regardless. The couple premium does not. The legal instruments are the door. The premium is what is on the other side of it.
Tier 1 — The standard: married couple raising their own shared biological children.
This is the benchmark, and it must be named as such. Conditions: legal marriage; at least one shared biological child in the household. The household receives the full couple premium, the direct tax benefits (joint filing, spousal allowances where applicable, inheritance treatment), the child benefit, the child-contingent supplements, and priority access to childcare and housing support. Time-unlimited, because the configuration does not become less valuable with duration.
Rationale: this is the first preference the introduction named, and every other tier is measured against it. The common-child stabilizer is present from the first day, the commitment filter has been exercised, the structural risk is the lowest of any configuration, and the fiscal externality is being generated in its cleanest form. A married couple with no children receives the child benefit the moment a child arrives, and the premium attaches at the same moment; before that there is no tier to occupy, because there is no externality to purchase. This is a timing rule, not a gap in the taxonomy.
Tier 2 — Reproductive committed stepfamily: reaches the Tier 1 standard.
Conditions: legal marriage; a shared child in common, or both partners under the reproductive-age threshold and within the plausible window for one (a defined duration, e.g., the first five years of the marriage); stepchildren present in the household.
Rationale: this is the only stepfamily configuration that produces a new net contributor while simultaneously stabilizing the children who already exist. It reaches the same standard as Tier 1—the same premium, the same tax treatment, the same child-contingent benefits—because on the dimensions the state is purchasing, it is doing the same work: a new taxpayer plus a stable second adult for existing children. It is a stepfamily, not the benchmark, and it is named as second-best in origin. But it converges on the benchmark’s function, so it converges on the benchmark’s subsidy. It also receives the stepfamily supplement and priority access to the childcare, housing, and counseling supports that make the configuration durable.
Tier 3 — Non-reproductive committed stepfamily, husband clearly out of his reproductive window: the full premium, time-unlimited.
Conditions: legal marriage; stepchildren present in the household; no shared child; the husband above a defined age threshold (a policy parameter—conservatively 55, or 50 with a documented fertility barrier), such that the shared-child stabilizer is structurally unavailable rather than merely delayed.
Rationale: the state is not buying a producer. It is buying a stabilizer—and it is buying the lowest-risk stabilizer in the taxonomy. The commitment filter is doing all of the work, the churn risk is minimal, and the fiscal downside that hangs over the young stepfather (a child he will not raise) is structurally absent. This is the configuration where the state gets exactly what it wants to buy—a stable second adult in a household raising the next generation—with the fewest moving parts. It should pay for it accordingly, and it should not impose a time limit, because the value of the arrangement does not expire. The children grow up, the household persists, and the fiscal benefit accrues for as long as the arrangement holds.
Tier 4 — Committed stepfamily, both partners in the reproductive window, no shared child: a time-limited, probationary premium.
Conditions: legal marriage; stepchildren present; both partners under the age threshold; no shared child after a defined period (e.g., three to five years).
Rationale: this is the highest-risk configuration in the taxonomy. It is the one most likely to be a “stage” rather than a commitment, the one in which the common-child stabilizer has not materialized and may never materialize, and the one in which the churn risk and the Cinderella risk are both at their peak—young stepfathers, young stepchildren, unformalized reproductive intentions. The state should support it, because the stepchildren still benefit from a second adult, but it should not support it indefinitely at the Tier 2 rate. The premium should be time-limited, renewed only on evidence of persistence—either the arrival of a shared child (which promotes the household to Tier 2) or the marriage surviving past the review (which is itself evidence that the commitment filter worked).
The cross-cutting instruments, applied at every tier above zero:
- Remove the couple penalty. Calculate means-tested benefits on individual income, or on household income with a generous couple allowance that ensures partnering never reduces household income. This is the precondition for everything else: it stops the benefit system from actively pushing couples apart.
- Add the direct couple premium. Pay it to formalized households that include a child or stepchild, at the tiered rates above. This is the positive incentive: it makes commitment measurably better off than cohabitation, so the couple is choosing with the state’s weight on the scale rather than against it.
- Condition the child benefit on the child, and the couple premium on the household structure. Pay the child benefit to the child’s household regardless of marital status. Pay the couple premium only where the commitment has been formalized, the household includes a child or stepchild, and the reproductive-window condition is satisfied.
- Recognize the stepfather legally and quickly. A fast, cheap adoption and guardianship pathway. The state should be paying for the second committed adult, not pretending the role does not exist.
- Fund the parenting support program as a condition of Tier 2 and Tier 4. The Building Strong Families finding—that pushing the least-committed couples into commitment made some outcomes worse—is a warning that the soft-skills component cannot be assumed. The program should be mandatory for the tiers where the risk is highest and voluntary for Tier 3, where the risk is lowest.
- Pilot and evaluate. The evidence base for subsidy design is thin, the Nordic experience suggests that legal recognition can close the cohabitation stability gap, and the BSF evidence suggests that coercion can backfire. The honest policy is a pilot with a proper evaluation, not a national rollout on an untested theory.
The principle that unifies the tiers. The state is not paying for marriage as such. It is paying for the next generation of taxpayers, and it is paying for the household structures that produce and stabilize them. Tier 1—the married couple with their own shared children—is the benchmark, the first preference, and the first claim on the budget. Tier 2—the reproductive committed stepfamily—does the same work and therefore receives the same standard. Tier 3 stabilizes the children who already exist at the lowest possible risk, and is paid for indefinitely because its value does not expire. Tier 4 may become either Tier 1 or Tier 3 and may become neither, and is supported on probation until it demonstrates which. And Tier 0, the unformalized cohabitation, however loving, has declined to make the commitment that lets the state distinguish it from the churn.
Part Eight: The Honest Bottom Line
The argument that the state should subsidize single mothers is correct, but it is already happening, and the subsidies are unconditional. The argument that it should subsidize the men who marry them is directionally correct, but the mechanism has to be the removal of the couple penalty, the direct couple premium, and the recognition of the second committed adult, not a bonus for the certificate alone. The argument that it should not sponsor mere cohabitation is defensible on actuarial grounds, but the strongest version of it is not a penalty on cohabitation—it is a refusal to extend the couple premium to a relationship form that is, on average, less stable, while offering cohabitants the legal instruments to make it more so. And the argument that the subsidy should be tiered by the reproductive window is the design variable that turns a slogan into a budget line.
The priority ordering must not be lost in the detail. The first claim on the state’s family budget is the freshly married couple raising their own shared children—Tier 1, the standard. That is the configuration the policy exists to encourage, and every subsidy discussed here is a second-best measure aimed at the households that configuration did not produce. The reproductive committed stepfamily is the one stepfamily form that reaches the same standard, because it does the same work. Stepfamilies are not the ideal. They are the reality, and the state has to fund them in their most stable and safest forms because the alternative is not the ideal—it is the absence of a second committed adult, at scale, in the households raising the next generation.
Children are a public good. The state should pay for them. But it should pay for the family form that produces them best, and it should stop penalizing the one it claims to want. The reproductive-window tiering is how that principle becomes a budget line.
The evidence is not clean. The selection problem is real. The Cinderella effect is real and the risk it represents is not dissolved by the commitment filter—only reduced and made visible. The marriage-promotion evaluations are disappointing. None of that changes the first-order recommendation, which is the one that survives every objection: stop taxing the thing you say you want, and start paying for it directly. Everything else is a pilot.
References
Adam, S., & Brewer, M. (2004). The “couple penalty” in the tax and benefit system. Institute for Fiscal Studies Briefing Note.
Anderson, K. G., Kaplan, H., & Lancaster, J. (1999). Paternal care by genetic fathers and stepfathers I: Reports from Albuquerque men. Evolution and Human Behavior, *20*(6), 405–431.
Antonovics, K., & Town, R. (2004). Are all the good men married? Uncovering the sources of the marital wage premium. American Economic Review, *94*(2), 317–321.
Brown, S. L. (2004). Family structure and child well-being: The significance of parental cohabitation. Journal of Marriage and Family, *66*(2), 351–367.
Chetty, R., Hendren, N., Jones, M. R., & Porter, S. R. (2018). Race and economic opportunity in the United States: An intergenerational perspective. Quarterly Journal of Economics, *135*(2), 711–783.
Daly, M., & Wilson, M. (1996). Violence against stepchildren. Current Directions in Psychological Science, *5*(3), 77–81.
Dion, M. R. (2005). Healthy marriage programs: Learning what works. The Future of Children, *15*(2), 139–156.
Eurostat. (2025). Children at risk of poverty or social exclusion in the EU.
Hsueh, J., Alderson, D. P., Lundquist, E., Michalopoulos, C., Gubits, D., Fein, D., & Knox, V. (2012). The Supporting Healthy Marriage Evaluation: Early impacts on low-income families. MDRC.
Kong, A., Frigge, M. L., Masson, G., Besenbacher, S., Sulem, P., Magnusson, G., … & Stefansson, K. (2012). Rate of de novo mutations and the importance of father’s age to disease risk. Nature, *488*(7412), 471–475.
Korenman, S., & Neumark, D. (1991). Does marriage really make men more productive? Journal of Human Resources, *26*(2), 282–307.
Lundberg, S., Pollak, R. A., & Stearns, J. (2016). Family inequality: Diverging patterns in marriage, cohabitation, and childbearing. Journal of Economic Perspectives, *30*(2), 79–102.
Manning, W. D., & Brown, S. L. (2015). Children’s economic well-being in married and cohabiting parent families. Journal of Marriage and Family, *77*(2), 345–362.
McLanahan, S., & Sandefur, G. (1994). Growing Up with a Single Parent: What Hurts, What Helps. Harvard University Press.
McLanahan, S., Tach, L., & Schneider, D. (2013). The causal effects of father absence. Annual Review of Sociology, *39*, 399–427.
Noack, T., & Wiik, K. A. (2008). Cohabitation and marriage in Norway: A comparison of family stability. In People, Population Change and Policies. Springer.
Teachman, J. (2002). Stability across cohorts in divorce risk factors. Demography, *39*(2), 331–351.
Wilcox, W. B. (2020). The Marriage Divide: How and Why Working-Class Families Are More Fragile Today. American Enterprise Institute.
Wood, R. G., McConnell, S., Moore, Q., Clarkwest, A., & Hsueh, J. (2012). The Building Strong Families Project: Effects of a relationship education program on unmarried parents. MDRC.