In 1973, Coale suggested that one of three preconditions for sustained fertility transition is that fertility control must be within “the calculus of conscious choice” (Coale 1973). Since then, the dominant approach of understanding reproductive decision-making and subsequent reproductive outcome has been intention-based. As introduced in Conceptualising fertility decision-making and behaviours , the behavioural models, such as theory of planned behaviour (TPB) (Ajzen 1985) and traits-desires-intentions-behaviour framework (Miller et al. 2004), have been applied in some surveys. In recent years, the 2011 Vienna Yearbook of Population Research featured studies which applied TPB to investigate reproductive decision-making process in Europe. Moreover, the largest source of health and pregnancy data globally are the Demographic Health Surveys, which ask preferences and intention of childbearing to all eligible women and men.

Although these intention-based approaches have made significant contributions in understanding reproductive preferences and behaviours, anthropological and social psychology literatures have raised concerns about the approach (Bachrach and Morgan 2013, Johnson-Hanks 2005). Couples may not have explicit plan of action in regard to reproduction. For instance, a study from the United States revealed that contraceptive use does not often corresponds to stated intention (Trussell et al. 1999). This may be a consequence of lack of perceived behavioural control, but some studies suggested that the behaviours may be habitual or reflection of underlying intention (Sable 1999).

Faced with an uncertain future, some women and couples have uncertain fertility intentions (Morgan 1981, Morgan 1982). The scepticism towards the meaning of fertility preferences and intention has been profound in sub-Saharan Africa. Timæus and Moultrie posit that fertility decline in sub-Saharan Africa is driven not by parity-specific limitation behaviours but by lengthening birth intervals, suggesting that women may postpone childbearing without clear decision over whether they want to have another child and when to have next pregnancy (Timæus and Moultrie 2008). A woman may not have decided whether she wanted any more children when she became pregnant. But the pervasive dichotomous question regarding fertility intention of a recent birth does not allow the woman to report the uncertainty.

However, several studies show strong predictive power. During 2 or 3-years follow-up, about 25-30% of women who wanted no more children gave birth, whereas about 60% of women who wanted another child had a live birth in Morocco, Egypt and Pakistan (Westoff and Bankole 1998, Casterline et al. 2003, Jain et al. 2014)

Qualitative and quantitative studies have shown that individuals may have ambivalent, contradictory or underspecified attitudes towards pregnancy. Women also expressed different degrees of strength of either trying to become pregnant or avoiding pregnancy.

A qualitative study in the United Kingdom found that intending to become pregnant and stopping contraception were not sufficient criteria to apply the intention into action in themselves; partner agreement and reaching the right time in terms of lifestyle/life stage were also necessary (Barrett and Wellings 2002). A quantitative study used the responses to the question "If you find you are pregnant in the next few weeks, would it be a big problem/would you be happy?" in the DHS from 3 African countries, and showed that 20-40% of women in Burkina Faso, Kenya and Ghana wanted to cease childbearing, but reported that it would be a small problem if they find that they are pregnant in the next few weeks (Speizer 2006). In a study in the USA, of pregnancies classified as contraceptive failure only 68% were unintended pregnancies and just 59% of these women reported that they were unhappy about the pregnancy (Trussell et al. 1999). Another study in the US showed that the parity progression ratio differs by level of certainty attached to fertility intention (Westoff and Ryder 1977).

A high degree of instability of fertility preferences among young women was suggested by studies in the United States, Ghana and Malawi (Kodzi et al. 2010, Sennott and Yeatman 2012, Rocca et al. 2010). Johnson-Hanks argues that the educated young Cameroonian women in her study adamantly do not provide a fixed preferences of childbearing because the future is uncertain and unpredictable, thus women need to stay adaptable (Johnson-Hanks 2002, Johnson-Hanks 2005).

In contrast, recent studies show high stability among married women who want to cease childbearing in Egypt and Malawi over 2-3 years at individual level (Casterline et al. 2003).

Fertility preferences may be not only be fluid but also sequential. Fertility preferences and intentions are likely to evolve systematically with age and existing family size, and respond to changes in people’s circumstances. Couples may strengthen certainty of their preferences or adjust their preferences after having the first child (Morgan 1981).

As discussed in Retrospective fertility preferences , There are substantial inconsistencies in the reporting of whether births were wanted or unwanted before and after conception of the birth These inconsistent reported fertility intentions are considered the result of rationalisation, but may also reflect a true change between a more abstract preference before the occurrence of childbirth and a more emotional reaction after the event (Cleland in press).