Mate choice can have important consequences for resulting offspring across species (Hamilton & Zuk, 1982; Jennions & Petrie, 1997). Less preferred or unattractive mates may result in offspring with lower viability, whereas preferred or attractive mates may result in offspring with increased fitness due to genetic factors (e.g., attractiveness is an honest signal of immune function), direct benefits (e.g., preferred partner has a territory with plentiful resources or provides parental investment), or some combination of both (Kirkpatrick & Ryan, 1991; Møller & Jennions, 2001; Petrie, 1994).

There are two hypotheses that make predictions regarding attractiveness (or preference) of one's mate and investment in subsequent offspring. The first is the differential allocation (DA) hypothesis (Burley, 1986) which argues that individuals should invest more in the offspring of attractive mates. Originally, this hypothesis proposed that individuals who reproduce with high quality mates should be willing to incur more parental costs than their mate to obtain or maintain attractive partnerships. In other words, an attractive mate may restrict their own parental investment which is made up for by their partner (Burley, 1986). In a study examining parental investment in zebra finches, Burley (1988) found attractive birds incurred smaller parental expenditures (i.e., they restricted their parental investment) whereas unattractive birds had larger parental expenditures (i.e., they increased their parental investment). While this hypothesis initially focused on species with biparental care, it has broadened to include species without biparental care (Ratikainen & Kokko, 2010). In Sheldon’s (2000) review of the differential allocation hypothesis, the “logic” is explained as allocating investment towards the offspring that provide the best return (like a bank with different investment rates). In this framing, both males and females may alter their investment related to the attractiveness of their mate regardless of their mate's parental investment. It has been debated whether the DA hypothesis as described by Sheldon (2000) is explicit in the direction of the prediction – should differential allocation always correspond to increased investment for offspring of attractive mates (Ratikainen & Kokko, 2010)? Following advice of Ratikainen and Kokko (2010) we use the term positive DA to refer to the positive correlation between mate attractiveness and parental investment. Support for positive DA has been found in a variety of species including birds, insects, house mice, and waterfrogs (Sheldon, 2000).

A second hypothesis, the Reproductive Compensation (RC) hypothesis, focuses on how individuals may try to compensate for mating with a less preferred partner by increasing investment (Gowaty et al., 2007). Also described as negative DA (Ratikainen & Kokko, 2010), this hypothesis proposes that individuals who mate under constraint and have offspring with lowered viability may compensate in a variety of ways, including reproductive effort (e.g., number of sperm ejaculated, eggs laid, or offspring born) and parental investment (e.g., size of egg laid) to compensate for lowered offspring viability (Gowaty, 2008). Theoretically, when pathogen evolution is rapid, complimentary immune alleles may increase offspring viability. If an individual is constrained to mate with a non-preferred partner (e.g., an individual with similar immune alleles) resulting in lowered offspring viability, an individual may compensate by increasing the number of offspring and in turn the variation expressed among them to improve the odds of survival to reproductive age for some (Gowaty et al., 2007). There are costs to compensation which should be observable in post-compensation mortality rates (Gowaty, 2008; Gowaty et al., 2007). Support for the RC hypotheses has been found in zebra finches (Bolund, Schielzeth, & Forstmeier, 2009), collared flycatchers (Michl, Török, Péczely, Garamszegi, & Schwabl, 2005), house finches (Navara, Hill, & Mendonça, 2006), and pronghorn antelope (Byers & Waits, 2006).

While these hypotheses may at first appear to make opposite predictions, it is important to point out the differences in these hypotheses. The (positive) DA hypothesis focuses on consensus attractive mates, e.g., those that most individuals of the opposite sex find attractive, while the RC hypotheses explicitly looks at preferred partners who are unique to each individual (Gowaty, 2008). Further, interpreting the diverse empirical findings is complicated by a variety of factors. First, “attractive” or “preferred” are measured or inferred in different ways and may have different effects on offspring viability. It is not always known how the attractive or preferred trait corresponds to offspring success. Is the ‘attractive’ feature correlated with some inherent genetic quality (such as immune function) or simply a trait that is deemed attractive but not correlated with any fitness outcome (other than one's offspring would inherit qualities deemed ‘attractive’). The mechanism by which preferred or attractive individuals influence offspring success may alter how individuals decide to allocate investment. Second, investment in offspring can be measured in a variety of ways that tradeoff with each other. If we assume that there is some inherent quality-quantity tradeoff, then when parents invest more in offspring is it to improve their quality (which we might assume under a “compensation” model) (Ratikainen & Kokko, 2010) or it is to improve quantity – perhaps to increase genetic variability in hopes that some offspring will succeed (Gowaty, 2008). These caveats make interpreting findings from different studies of varying species difficult.

Many tests of these hypotheses have been conducted and while more have found support for the positive DA hypothesis, there are some that have supported the RC hypothesis and others that have found no effect or mixed results. In some cases, the same species has been used to support both hypotheses (e.g., the zebra finch (Bolund et al., 2009; Burley, 1986)). Some research that supports the positive DA hypothesis includes Pryke and Griffith (2009) who found that female Goudian finches increased their provisioning efforts nearly twice as much when breeding with compatible males compared to mating with incompatible males (those of a different genotype). The authors suggest that reproductive compensation is more likely to occur in situations where low-genetic quality offspring, if provided with adequate resources during development, may have the same potential survivorship and reproductive success as high-quality offspring. Harris and Uller (2009) developed a mathematical model to explicitly compare these hypotheses. They found that the most common strategy for females under a variety of parameters would be increased investment when mating with mates that produce higher viability offspring. When mating with low-quality mates, increased reproductive investment was predicted to occur only when the baseline offspring survivorship was relatively high or when the quality of future mates was expected to be low. The key to understanding various empirical findings may be to explicitly identify the association between parental investment and offspring fitness, the expected future mating opportunities, and the increased mortality risk related to compensation that may vary across species, individuals, and even within an individual's life.

To test the reproductive compensation (RC) and positive differential allocation (DA) hypothesis in humans, it is first necessary to consider what sorts of constraints on mate choice exist in human populations. One circumstance that may cause people to accept less preferred mates is that of arranged marriage, a unique human behavior where parents and sometimes other kin, either choose or influence the choice of someone else's marital partner (Batabyal, 2001; Dubbs, Buunk, & Taniguchi, 2013). Arranged marriage is a common form of marriage across many human populations historically (Apostolou, 2014; Buunk, Park, & Dubbs, 2008; Hasnain & Snopkowski, 2023). Apostolou (2007) examined the prevalence of arranged marriages in 190 hunter-gatherer societies around the world and found that in 94% of foraging societies, parental arrangement, close-kin arrangement, and courtship with parental approval, are the predominant forms of marriage. A study examining sixteen historical societies found that marriages arranged by parents (principally fathers) were the most frequent form of marriage (Apostolou, 2012). A phylogenetic reconstruction of the ancestral marriage pattern using mitochondrial DNA sequences of extant hunter-gatherers suggests that arranged marriage may have a deep origin, possibly >50,000 years ago (Walker, Hill, Flinn, & Ellsworth, 2011). These results show that arranged marriage is an important feature of human marriage historically.

Parental control of marriage arrangements would not be considered a constraint on mate choice if the interests of both parents and offspring are the same. However, conflict may exist between parents and offspring due to differences in genetic relatedness (Trivers, 1974). Since parents are more related to themselves than their children and have the same probability of sharing a gene through descent with all offspring, they may prefer a partner for their children who will improve (or at a minimum maintain) the family's status, wealth, and reputation (Apostolou, 2007; Buunk et al., 2008). If parents choose partners, it is predicted that they focus on the success of the entire family and frequently view marriage as the joining of families (not individuals). In contrast, when individuals choose their own partners, there may be a greater focus on a partner who provides benefits to their spouse and children over the benefits provided to the larger family group (Apostolou, 2011; Buunk et al., 2008; Buunk & Solano, 2010).

Other research has instead focused on the benefits of arranged marriage for children, seeing it as a cohesive family strategy (e.g., Dasgupta, 2009; Hasnain & Snopkowski, 2023; Mace & Mace, 1960). Parents have more experience with marriage and may be better able to identify whether a person will be a good partner for their child. Further, there may be benefits that children gain if they marry a partner chosen by their parents, such as greater social or financial support after marriage (Agey, Crippen, Wells, & Upreti, 2023).

A variety of empirical studies, usually using surveys, find that the traits offspring seek in a mate are not the same traits that parents value in an in-law (Agey, Morris, Chandy, & Gaulin, 2021; Apostolou, 2011, Apostolou, 2017; Bovet, Raiber, Ren, Wang, & Seabright, 2018; Buunk et al., 2008; Buunk & Solano, 2010; Dubbs et al., 2013; Dubbs & Buunk, 2010; Gangestad & Simpson, 2000; van den Berg, Fawcett, Buunk, & Weissing, 2013). For instance, surveys have shown that individuals prefer personality traits such as having an exciting personality, creativity, or sense of humor more than parents do (Apostolou, 2011; Buunk et al., 2008; Dubbs et al., 2013; Dubbs & Buunk, 2010; Fugère, Doucette, Chabot, & Cousins, 2017; Perilloux, Fleischman, & Buss, 2011). In contrast, parents prefer traits that indicate parental investment and cooperation, and prefer potential mates for their offspring that have the same religious, ethnic and social class (Apostolou, 2008, Apostolou, 2011; Buunk et al., 2008; Buunk & Solano, 2010; Dubbs et al., 2013; Dubbs & Buunk, 2010; Perilloux et al., 2011). These data suggest that there is conflict between children and their parents over prospective mates and that under conditions of arranged marriage, children may marry partners they view as less preferred.

Little is known about how these different preferences play out when comparing people who choose their own marriage partners to those whose marriage partners are chosen by others, as most work has utilized surveys and hypothetical examples. A recent study examining reproductive success (number of living offspring) in three indigenous populations found that couples in arranged marriages had similar reproductive success to those in self-choice marriages (Sorokowski et al., 2017). This result is counter to much of the previous work that has been done suggesting that free mate choice is important for fitness outcomes. Our study will explore these questions in more detail, examining a wider variety of possible differences in both reproductive output and parental investment.

Parental investment was defined by Trivers (1972) as, “any investment by the parent in an individual offspring that increases the offspring's chance of surviving (and hence reproductive success) at the cost of the parent's ability to invest in other offspring”. For this study, we focus on maternal investment, since women report on investment in their offspring and there is little information about paternal investment. If a woman is married to a less desirable partner, it is possible that she will invest more in offspring (to compensate following the RC hypothesis) or invest less (following the positive DA hypothesis). This leads to our research question: Are there differences in maternal investment strategies between women whose marriages were arranged and those who chose their own partners? Given the differing predictions of the RC and the positive DA hypothesis, we hypothesize that women in arranged marriages will invest a different amount than those in self-choice marriages. If women in arranged marriages invest more than those in self-choice marriage, it would provide evidence for the RC hypothesis. In contrast, if women in self-choice marriages invest more than women in arranged marriages, it would provide evidence for the positive DA hypothesis.