Marriage used to be practically universal, but now persists as an institution for only some groups, while others choose non-marital fertility. This paper posits that if one role of marriage is to insure one partner's investment in children, then home-ownership can be seen as providing the necessary "collateral" for this contract. As easier divorce and paternity enforcement outside of marriage have reduced the relative strength of the marital contract, the division of assets, particularly the family home, post-separation has remained unique to marriage. We provide a model where husbands can "ante up" the marital home to elicit more optimal child investments, whose costs are born mostly by the mother, by reducing the chance of divorce while providing consumption insurance to their partner. This, in turn, increases the value of marriage for those able to access this collateralized version of the contract. The model predicts that individuals able to buy a home at the time of marriage will invest more in children and have greater labor specialization, while policy changes that eroded marriage's relative commitment value would have heterogenous effects by asset-holding, both of which appear to hold in US data.