Uncertainty – a condition with unknown probability distributions of future outcomes – represents an intrinsic characteristic of contemporary societies. The ideas of “risk society,” “reflexive modernity,” or “liquid modernity” describe a historical trend of the last decades in which uncertainty is a new feature of social change (Bauman, 2000; Beck, 1992; Giddens, 1991). More than a century ago, Karl Marx designated modernity itself as a novel era in which “all that is solid melts into air” (1848/2020, p. 475), but the increasing speed of technological change, the constant flows of financial capital across the globe, labor market reforms and, more recently, climate change and its social consequences have expanded the sources of uncertainty. These conditions of uncertainty affect private lives (Sennett, 1998) and family life courses (Kreyenfeld et al., 2012; Mills and Blossfeld, 2013). Embedded in such a contemporary scenario, the COVID-19 pandemic of 2020, and the responses to the outbreak, escalated the uncertainty at the core of the public debate and personal lives. Policy-makers, but also scientists, have no clear answers to the questions of how long the pandemic will last, and what the real consequences will be for public health, as well as for other social and economic outcomes. Especially in countries that implemented nationwide lockdowns, people started to feel insecure in their daily lives due to the risk of contagion, which also depended on others’ “safe” behavior. The possibility of losing one's job and/or having a reduced standard of living is a widespread renewed source of concern in response to a looming economic future that nobody can forecast, even in the short term. This additional condition of uncertainty can be seen as an ancillary outcome of globalization, as the rapid diffusion of the pandemic is also related to the large volume of exchanges and global interdependencies (Kaufmann, 2009).

Thus, while media reports speculated about a surge in “corona-babies” conceived during the pandemic and its related lockdowns, it seems plausible to expect an additional negative impact on family formation due to the increasing uncertainty about the future, at least in high-income countries (Aassve et al., 2020). In this article, we explore the consequences that the rising uncertainty induced by the COVID-19 pandemic may have for fertility intentions. The latter reflect the combined effects of desired fertility and situational constraints (Billari et al., 2009), and have been generally regarded as a fairly reliable predictor of behavior, provided that a time frame for their realization is set (Schoen et al., 1999; Westoff and Ryder, 1977).

The pandemic occurred within a context of demographic change in which fertility rates in many countries in Europe and the US had been declining during much of the 2010s. The underlying nature of this decline is still a conundrum for demographers and sociologists. Fertility decreased dramatically both in already low-fertility countries of Southern Europe that were severely affected by the economic and social consequences of the Great Recession; and in Nordic countries such as Norway and Finland, which experienced an almost immediate recovery of economic growth, and where the institutional context continued to provide a more favorable environment for childbearing. There have been several empirical attempts to understand the reasons underlying the fertility decline after 2008, but even studies that simultaneously included several indicators of economic conditions, such as the unemployment rate, foreclosure rates and the cost of public debt, were not able to fully explain the relatively homogeneous fertility contractions that Western societies are currently facing (e.g., Comolli, 2017; Goldstein et al., 2013; Matysiak et al., 2020; Schneider, 2015). This is because, we posit here, objective indicators of individuals’ employment and economic conditions subsume the “statistical shadow of the past” (Davidson, 2010, p. 17), which tells us little about the uncertain future that people experienced during the crisis.

We argue that research on fertility decisions in uncertain times needs to partly shift its perspective, recognizing that uncertainty is a forward-looking notion. The study of the influence of the future in decision-making processes has a long tradition in the social sciences, and there has been renewed interest in this topic in recent years. We rely on the Narrative Framework for the analysis of fertility intentions (Vignoli et al., 2020a, 2020b), which is based on recent developments in economic sociology on decision-making under conditions of uncertainty (Beckert, 2016; Beckert and Bronk, 2018). This future-oriented framework represents a novelty in the study of fertility decision-making processes, and it is obviously applicable to the analysis of the consequences of the COVID-19 emergency. The pandemic, and the related lockdowns, indeed represent a situation in which the ordinary temporal orientation is suspended: the degree of “clarity” with which the future is imagined is reduced, and the future horizon is “contracting” because forecasting is more difficult than it was before (Mische, 2009, p. 700). During the pandemic, the expected future has been shaped by individuals’ direct exposure to the SARS-CoV-2 virus – i.e. by their own or their close relatives’ exposure to contagion and subsequent social isolation, hospitalization, or even death – and the economic consequences of the pandemic, such as job loss or temporary inactivity due to a lockdown. However, for the majority of the population who have not experienced severe health and economic consequences due to the pandemic, expectations are shaped by the spread of shared narratives of an uncertain future (Vignoli et al., 2020b), especially those channeled by the media and related to the diffusion of the virus, government restrictions and the scientific debate about when the pandemic will be over (Egidi and Manfredi, 2021).

In Italy, our case study, the COVID-19 disaster has created an enormous uncertainty shock, which, without operationalizing the whole Narrative Framework, allows us to make a first explorative attempt to test empirically the influence of the “shadow of the future” (Bernardi et al., 2019, p. 4; Huinink and Kohli, 2014, p. 1303) on fertility intentions under conditions of uncertainty. Italy was affected more strongly by the 2008 recession and by the public debt crisis than many other industrialized countries. While Italy had a total fertility rate of 1.46 in 2010, it reverted to a lowest-low fertility regime in 2019, with a total fertility rate of 1.29, which was the lowest in Europe. On top of that, Italy was the first country in the Western world to be severely hit by the COVID-19 pandemic, with Italians experiencing the longest complete and nationwide lockdown, which started on March 9 and ended on May 4, 2020. Because of COVID-induced uncertainty, subjective perceptions and personal narratives of the future – which are also influenced by media-channeled shared narratives – may gain the upper hand over the “shadow of the past” for fertility intentions. What has happened to childbearing plans during this unexpected period of uncertainty about the future? Have fertility intentions been negatively affected by the pandemic? Can the impact of the pandemic be explained by people's objective exposure to the virus and its related socioeconomic consequences, or is it better grasped by exploring people's rising uncertainty about the future, which has also been spread by the media?

To address these questions, we make use of unique data that we collected during the spring 2020 lockdown on a sample of Italians of reproductive ages. We measured individuals’ expectations concerning the duration of the pandemic emergency and family imaginaries, contrasting their effects on fertility intentions during the lockdown with those of past experiences, and the individuals’ objective levels of exposure to the pandemic and its socioeconomic consequences. In addition, we provide empirical evidence of a causal impact of shared narratives of the future on fertility intentions by making use of online experimentation, an innovative approach to the study of the impact of the future in decision-making processes. Our experiment simulates a “real” exposure of respondents to a new media narrative. Respondents were randomly assigned to read different mock news bulletins concerning the expected end of the pandemic emergency in Italy, each of which presented a different expected duration of the crisis before a return to normality. We then compared their post-treatment and pre-treatment fertility intentions.

Much of the literature on fertility is based on the study of the social determinants of fertility, which mainly accounts for the influence of what already happened in previous stages of the life course, and thus considers factors such as educational attainment, previous (un)employment episodes and partnership histories (Barbieri et al., 2015; Busetta et al., 2019; Kreyenfeld et al., 2012; Vignoli et al., 2020c). These experiences are shaped during socialization and by personal predispositions, like risk aversion or time discounting preferences, which may also have a direct influence on fertility choices (Bellani et al., 2021; Schmidt, 2008). Psychological predispositions, cumulative past experiences over the life course and the present socioeconomic status are the standard elements usually identified as determinants of fertility intentions and behaviors (Busetta et al., 2019; Dantis and Rizzi, 2020; Mills and Blossfeld, 2013; Vignoli et al., 2012), and are aspects that need to be controlled for in any empirical model of fertility intentions. However, this “driven-by-the-past framework” (Seligman et al., 2013, p. 127) makes agency and choice difficult to understand, as fertility decision-making is a complex process that is influenced, but not determined, by past experiences.

The influence of the future on the course of action is difficult to conceptualize and operationalize in empirical research. The pragmatist philosophical tradition devoted special attention to the role of the future in the course of action. Following this approach, we can posit that future expectations are not just determined by pre-existing cognitive schemas or typification (Schütz, 1962), but are deeply imbued with imaginative capacity (Dewey, 1922/1930; Mead, 1932/2002). Of course, not all human actions are the result of deliberative thinking. According to Dewey, the ordinary course of action is an unreflective flow of activities in which “habits do all the perceiving, recalling, judging, conceiving and reasoning that is done” (1922/1930, p. 177). However, the ordinary, unconscious course of action can be interrupted by the emergence of conflict between “different habits, or by the release of impulses,” or when the actor is confronted with a “new and surprising situation” (Beckert, 2016, p. 54) in which the expected outcome of the ordinary routine no longer seems to apply. At this point, people experience uncertainty about the future, and the deliberative process emerges, as the situation requires a (new) judgement. In a situation of uncertainty, past experiences and expectations come into play in an imaginative “dialogue” in which “competing possible lines of action” are considered because “deliberation is an experiment in finding out what the various lines of possible action are really like” (Dewey, 1922/1930, p. 190). The influence of the future in fertility plans has previously been considered in the New Home Economics approach (Becker, 1981) and the psychological Theory of Planned Behavior (Ajzen, 1991; Ajzen and Klobas, 2013). However, these approaches undermined the role of human agency, and failed to provide a systematic framework to account for its importance in explaining fertility dynamics.¹The Narrative Framework – presented in Figure 1 – identifies the key elements that are involved in this future-oriented deliberative process: expectations, imaginaries and personal narratives of the future, which define what can be referred to as the “shadow of the future” that influences the decision-making process.

The figure represents the different steps of a stylized decision-making process. Each element of the framework stems from the previous ones, but can also exceed them and have an independent effect on fertility decision-making. Although these elements are not meant to follow any strict order in actual decision-making, from an analytical point of view, it is useful to start considering psychological predispositions, past experiences (B) and current (socioeconomic) status (C), which represent a set of opportunities and constraints for childbearing plans. A recent stream of literature has introduced personal perceptions (D) of past and current experiences as a way to introduce agency into empirical models of fertility intentions and behavior, and they account for the fact that people may react very differently to the same objective experiences and economic conditions (Kreyenfeld, 2010, 2015). Individuals’ perceptions have been found to play an independent role net of objective indicators of individuals’ past and current labor market situations (e.g., Bhaumik and Nugent, 2011; Fahlén and Oláh, 2018), and to moderate the impact of these indicators on fertility intentions (Vignoli et al., 2020d) and behavior (Kreyenfeld, 2015). This approach is in line with the Thomas theorem, which posits that the interpretation of a situation causes the action (Thomas and Thomas, 1928). While perceptions of insecurity are related to individuals’ current circumstances, they obviously refer to possible future events or threats. However, individuals’ subjective evaluations of the (in)security of their current conditions only implicitly entail a reference to the future. In this sense, perceptions are somewhat “in between” the shadow of the past and the shadow of the future. To address theoretically and empirically the role of uncertainty in fertility decisions, a conceptualization that explicitly acknowledges its forward-looking nature is needed.

In Figure 1, expectations (E) are the first step into the shadow of the future, as they represent what people expect will happen in the future based on the available information. Although expectations may arise from past experiences, they are often connected to a shadow of the future, and, thus, become an independent source of agency. For instance, working with a fixed-term contract may not negatively affect an individual's fertility intentions if he/she believes that economic growth will be strong or that permanent employment opportunities will increase.

However, expectations do not account for the full influence of the future on the course of action, as imaginaries (F) may shape and deviate from an expected future. Imagination is the capacity to place oneself in one or more imagined situations, while also hypothesizing alternative courses of action and their effects. But imagination, more radically, also allows individuals to imagine a possible future that cannot be deduced from their present circumstances. Personal imaginaries may be easily influenced by social norms and relevant others’ opinions, but may also deviate from them, and can thus move the decision-making process in a different direction. For instance, while the two-child norm is widespread in wealthy countries (Sobotka and Beaujouan, 2014), a personal family imaginary may revolve around a one-child or even a childless family. Imaginaries constitute a less abstract point of reference than social norms because they represent wishful (or fearful) projections into the future, which arise from the capacity of human agency to shift away from the expected course of action (Emirbayer and Mische, 1998). Imaginaries may play a crucial role in decision-making, and especially when making decisions that are likely to have complex and long-term outcomes, as in the case of fertility decisions. Long-term outcomes cannot be forecast, and/or each possible future may involve both positive and negative expectations. A normative orientation related to personal imaginaries (“How I would (not) like the future to be”) may come into play and orient deliberation, shedding a special light on the different pros and cons implied by the available options, and thus help to orient the decision. A psychological mechanism compatible with the importance of imaginaries can be found in the “affective forecasting theory,” which posits that people base their decisions on affective forecasts; i.e., on their predictions about their own emotional reactions to future events (Wilson and Gilbert, 2003). Demographic research has shown that the happiness of parents-to-be increases before childbirth (Myrskylä and Margolis, 2014), and that the anticipation of an increase (or a decrease) in one's own happiness from having a(nother) child may influence the decision to have the child at all (Billari, 2009). According to the Narrative Framework, family imaginaries represent the source of the “expected happiness” from childbearing, and, thus, have a dual effect on the cognitive process of deliberation. First, they provide a frame in which the current status (C) and the perception of the current status (D) are interpreted and evaluated, and which cannot be reduced to expectations (E), such as whether or not the (un)employment situation influences childbearing plans. Second, they represent an independent source of a conscious desire for a change in the future; i.e. they may provide a life goal, irrespective of the shadow of the past and (more or less plausible) expectations.

Personal narratives of the future (G) reflect the contingent plan for reaching the goals set by the imaginaries. As Figure 1 suggests, narratives do not just add an additional element to the framework; rather, they represent the less abstract level of the decision-making process, in which the shadow of the past, expectations and imaginaries find their proper places, and, at this level, influence fertility intentions (Vignoli et al., 2020a). All previous elements in the framework are selected, interpreted and included in a personal narrative of the future, which also entails a hypothetical course of action and the causal interconnection of these elements.

Life course decisions like fertility decisions necessarily involve a conscious narrative of the future, which embodies the causal path that people deem necessary to reach their imagined goal. Expectations, imaginaries and narratives of the future might facilitate or inhibit fertility in conditions of uncertainty. They might foster fertility in line with the socio-psychological uncertainty reduction framework from Friedman et al. [1994], who argued that more economically vulnerable women may respond to uncertain life prospects by choosing to become a mother, which gives meaning and stability to their lives. However, empirical evidence suggests that people, and especially young people, usually build their personal narratives to act in accordance with a condition of economic uncertainty in order to avert risk (McDonald, 2002; Schmidt, 2008). Young adults tend to postpone making long-term, binding decisions, such as decisions about marriage and childbearing, until they become more settled in the labor market (Mills and Blossfeld, 2013), and such tendencies may be particularly strong in a country like Italy, which is characterized by the “postponement syndrome” (Livi Bacci, 2001), and where the perceived economic preconditions for family formation seem very high (Vignoli et al., 2020b). Moreover, a family imaginary may revolve around the desire to remain childless. For these individuals, all of the previous elements of the Narrative Framework play only a marginal role in defining the personal narrative that influences their fertility decisions.

Individual actors are not the only “authors” of their own narratives of the future, as they are influenced by factors external to them, in the form of context and shared narratives (A). The “context” usually considered in comparative analyses of fertility is related to the institutional setting, prevailing values and long-term cultural continuities (Balbo et al., 2013). However, above and beyond the influence of these contextual factors, a last element of our Narrative Framework is represented by shared narratives (A): i.e. narratives of the future adopted by relevant others such as parents and peers, or conveyed by the media (Vignoli et al., 2020b). Parental pressure is likely to influence young people's family plans, especially in a “strong family” setting in which young adults tend to leave their family of origin relatively late, such as in Southern European countries (Billari, 2004). This is a situation that may not apply to other Western European countries, where the influence of peers may be more relevant (Di Giulio and Rosina, 2007; Guetto et al., 2016). However, in recent years, a major source of influence in globalized societies is the unprecedented access to press and new (social) media, which may shape individuals’ perceptions, expectations and imaginaries.

Media-channeled shared narratives have started to play a major role in spreading feelings of uncertainty about the future in recent decades, and the media coverage of economic issues has increased substantially, especially after the 2008 crisis, in both Europe and the US (Baker et al., 2016). This constant (over-)flow of information, which was further intensified after the start of the pandemic (Altig et al., 2020), is likely to exacerbate individuals’ feelings of uncertainty about the future because of the prevailing tone and angles of media reports (Alsem et al., 2008; Dräger, 2015). Schneider [2015] suggested that press coverage comes closer to measuring the sentiments that shape economic uncertainty and that affect fertility decisions than actual economic constraints. More recently, the European sovereign debt crisis has received considerable attention in the media, and the simplified narratives presented in the press have served as a multiplier of uncertainty, contributing to shrinking birth rates (Comolli and Vignoli, 2021).²

In this paper, we propose an application of the Narrative Framework outlined above to the study of fertility intentions during the COVID-19 pandemic outbreak, with an empirical focus on the first nationwide lockdown implemented in Italy. Although this empirical exercise does not aim to operationalize the whole theoretical scheme presented so far, two sets of hypotheses refer to the “shadow of the past” of COVID-induced uncertainty (Hypotheses H1a, H1b) and to the “shadow of the future” of pandemic uncertainty and personal family imaginaries (Hypotheses H3a, H3b). Given their liminal positioning at the intersection between the two “shadows” (Figure 1), a specific hypothesis (H2) refers to the role of perceptions related to personal and general situations. We also take into account the role of media shared narratives during the pandemic, and their possible causal impact, through an ad hoc experiment (Hypothesis H4).

The pandemic has exogenously exposed people of reproductive ages to a new environment characterized by a high level of uncertainty. This new, uncertain situation affects individuals through two main mechanisms: the health and economic consequences of the pandemic and related government restrictions (context) on the one hand, and exposure to the (social) media coverage of the pandemic (shared narratives) on the other.

The pandemic context has led to important changes in individuals’ objective status, first and foremost in terms of the potential health consequences of direct exposure to the SARS-CoV-2 virus. Apart from those individuals who needed hospitalization because they had more severe symptoms, people who tested positive for the virus had to face quarantine and social isolation, and even many untested people were put in quarantine because of a suspicion that they or their close relatives or acquaintances had been infected. The uncertainty created by this unexpected situation is likely to hinder childbearing plans, both directly and indirectly due to induced perceptions of insecurity, which leads to our first hypothesis:

Hypothesis 1a. Personal, close relatives’ or acquaintances’ exposure to the SARS-CoV-2 virus negatively affects pre-pandemic fertility intentions.

The economic consequences of the pandemic have been even more widespread. During the first lockdown, which started on March 9, 2020, Italians were prohibited from leaving their homes, except to engage in work activities deemed “essential,” to buy food, or in cases of utter emergency. The most fortunate workers – usually highly educated individuals employed in skilled jobs and in regular forms of dependent employment – shifted to working from home, and did not face a serious risk of being laid off or suffering earning losses, especially if they were public sector employees. However, the Italian National Institute of Statistics (ISTAT) estimated that approximately one-third of the total labor force were employed in economic sectors whose activities were suspended, especially in the private sector, in which almost half of firms were affected by government restrictions (ISTAT, 2020). This meant that millions of suspended employees had to rely on wage guarantee funds, which entailed a 35% average reduction of the usual salary; whereas many self-employed individuals, especially in the consumer services sector, had to temporarily interrupt their activities with limited or no earnings, apart from discretionary lump-sum transfers provided by the government to help them weather the emergency. Finally, many workers lost their jobs, although the exact numbers of such workers are difficult to estimate, especially if they were employed with temporary contracts or in the black economy. Restrictions were gradually loosened by the government up to the end of the lockdown on May 4, 2020. Many businesses, such as bars and restaurants, did not re-open until the beginning of June, and travel between Italian regions (NUTS-2) without any certified urgent reason was not allowed until June 3, 2020. In light of the available empirical evidence concerning the negative effects of unemployment and more unstable labor market conditions on fertility in Italy (Alderotti et al., 2021; Busetta et al., 2019), this objective situation of economic uncertainty leads to the next hypothesis:

Hypothesis 1b. The experiences of work suspension and/or job loss because of government restrictions negatively affect pre-pandemic fertility intentions.

The new context created by the pandemic and government restrictions influenced many aspects of individuals’ objective status, which, together with their past experiences and psychological predispositions, are part of the shadow of the past in the fertility decision-making process during the COVID-19 pandemic. However, the pandemic context also influenced more subjective states through exposure to media shared narratives. Indeed, during the lockdown, most people were informed about the diffusion of the pandemic and government decisions only by the (social) media, which focused heavily on trends in the diffusion of coronavirus infections. During this period, Italians gathered in front of their TVs every day at 6 p.m. to get the official updates concerning the daily numbers of hospitalized individuals and deaths from the Civil Protection's press conference. The daily news of coronavirus deaths and infections was a major source of concern for the population in this period. In the first half of April 2020, at the peak of the pandemic and in the middle of the lockdown period, Italy recorded an average of 700 COVID-related deaths per day. As of April 15, Italy had approximately 150,000 positive cases and 20,000 official deaths, which were the highest numbers in Europe at that time. It is important to note that these figures are difficult to compare across countries due to various kinds of biases in the identification of the exact cause of death (Odone et al., 2020).³Nonetheless, the awareness of these official numbers is likely to have shaped individuals’ perceptions of the seriousness of the emergency, and to have added a generalized feeling of insecurity to individuals’ baseline psychological predispositions to risk aversion, even among those who were not directly exposed to the objective socioeconomic consequences of the pandemic. This leads to our next hypothesis:

Hypothesis 2. Perceptions of insecurity across several life domains due to the pandemic negatively affect pre-pandemic fertility intentions.

The public media discourse in Italy was filled with heated debates among virologists, politicians and opinion leaders on issues such as “the exponential growth of infections,” the need for “measures to flatten the curve” and whether (and which types of) masks are useful to slow down the pandemic, often with contradictory messages (Ruiu, 2020). In line with our Narrative Framework, people were thus exposed to a high degree of uncertainty about their future, and had to form their own expectations about when the return to pre-pandemic conditions would occur. Expectations are the first element of the shadow of the future that have influenced fertility intentions during the pandemic:

Hypothesis 3a. The expectation that the return to pre-pandemic conditions will take a long time negatively affects pre-pandemic fertility intentions.

Taken together, people's current status, perceptions and expectations about the length of the emergency – all of which are influenced by the new, uncertain context created by the pandemic and the related media shared narratives – are expected to influence people's fertility intentions during the pandemic, net of past experiences and psychological predispositions. Personal family imaginaries related to the joy of parenthood are also expected to influence changes in fertility intentions, although they should be only marginally affected by the pandemic. Imaginaries do change during the life course, but since they usually incorporate a normative orientation toward the future, they are not likely to be modified in the short term (Kiley and Vaisey, 2020). The presence of family imaginaries and their influence on fertility intentions leads to our next hypothesis:

Hypothesis 3b. Individuals with a positive family imaginary are less negatively influenced by the pandemic, and are less likely to change their pre-pandemic childbearing plans.

For people who were not directly exposed to the health and economic consequences of the pandemic, the shared narratives spread by the media were the major sources of uncertainty. However, to effectively grasp the media effects, indicators of the media coverage of the pandemic, combined with measures of individuals’ exposure to those specific media contents, would be needed. Moreover, even if all of the necessary information was available, causality would remain difficult to ascertain. For these reasons, we adopted an experimental approach in which respondents were exogenously exposed to different scenarios regarding the expected end of the pandemic, mimicking a news report. We then asked the respondents about their (renewed) fertility intentions, in light of the expected duration of the emergency. Using this approach, we both provide additional evidence on the importance of media shared narratives and reinforce the claims about the causal role of the shadow of the future in fertility intentions. This experiment allows us to test our last hypothesis:

Hypothesis 4. Exposure to a new shared narrative that the return to pre-pandemic conditions will take a long time negatively influences fertility intentions; whereas exposure to a new shared narrative that the return to pre-pandemic conditions will happen quickly positively influences fertility intentions.

To sum up, if fertility decisions have been negatively influenced by the pandemic, we hypothesize that this effect cannot be explained by the objective exposure to the pandemic's health and economic consequences alone (Hypotheses H1a and H1b). In a context of amplified uncertainty about the future, perceptions (Hypotheses H2), and, in particular, expectations and imaginaries, may gain the upper hand (Hypotheses H3a and H3b). Expectations about the length of the pandemic spread by the media, which are simulated through our experiment, may also have a specific influence on fertility intentions (Hypothesis H4).

This first application of the Narrative Framework to the empirical analysis of changes in fertility intentions during the pandemic is carried out by means of survey data that we collected during the lockdown in Italy. These data operationalize all of the elements included in Figure 1, but do not include personal narratives of the future, which are more easily explored through qualitative interviewing. However, if relevant information regarding all of the elements involved is available in the survey data, it may be used as a proxy to grasp the effects of personal narratives (Vignoli et al., 2020b).

In this article, we argued that the explanation for individual decision-making in conditions of uncertainty – especially for long-term, binding decisions, such as fertility decisions – needs to provide a complex account of different temporal orientations. While the sociological and demographic literatures have widely demonstrated that fertility decisions are shaped by individuals’ previous life experiences and socioeconomic status – which we refer to as the “shadow of the past” – rising uncertainty about the future necessitates an analytical framework that explicitly acknowledges its forward-looking nature. Building on recent developments from economic sociology (Beckert, 2016; Beckert and Bronk, 2018), we relied on the Narrative Framework (Vignoli et al., 2020a, 2020b), which argues that personal narratives of the future, and their constitutive elements of expectations and imaginaries – which we refer to as the “shadow of the future” – represent crucial drivers for decision-making under conditions of uncertainty. Personal narratives of the future are not merely the result of psychological predispositions or idiosyncratic preferences, but are socially constructed in that they are shaped by “shared narratives” in the form of public images produced by the media and other powerful opinion formers.

We exploited the exogenous shock provided by the COVID-19 pandemic as a crucial occasion to test empirically some of the predictions of the Narrative Framework, without having any explicit aim to operationalize the whole theoretical schema. We argued that narratives of the future have become particularly important for fertility decision-making in contemporary societies, but also that the recent COVID-19 pandemic represented an enormous uncertainty shock that could have made the role of the shadow of the future particularly salient, over and above the effects of the shadow of the past. For instance, even scientists found it difficult to forecast the length and the consequences of the pandemic. Meanwhile, people needed to cope with leading a different daily life; form their own expectations about when conditions would return to “normal;” and, on that basis, formulate their life plans, including about family formation. Government restrictions that were imposed following the start of the pandemic may not have negatively influenced people's intentions to have a(nother) child if they anticipated a rapid return to pre-pandemic conditions, whereas these restrictions might have inhibited fertility if people had a darker view, and thus expected uncertainty to persist, irrespective of their current socioeconomic status and perceptions of it. At the same time, in such an unprecedented situation, in which expectations could not be based on any firm grounds, imaginaries may have played an additional, independent role.

Using unique data collected during the spring 2020 lockdown in Italy, the first Western country to experience widespread diffusion of the SARS-CoV-2 virus, we showed that objective indicators of individuals’ exposure to the health and economic consequences of the pandemic played a very limited role in re-shaping individuals’ fertility plans during the lockdown. This is not to say that the shadow of the past was not at all relevant: traditional predictors, such as age, sex, the number of children and the area of residence were significantly associated with changes in fertility intentions. However, people's subjective perceptions, also supported by media-channeled shared narratives, and especially their expectations and imaginaries, were found to be crucial moderators of their adaptation to a new context characterized by overwhelming uncertainty. Hence, taking the shadow of the future into account is a more effective way to understand the mechanisms through which the pandemic affected fertility intentions than simply measuring people's objective exposure to the virus and its related socioeconomic consequences. Our results suggest that when respondents expected that it would take a long time for their pre-pandemic life to return to normal, their pre-pandemic fertility intentions were reduced.

Importantly, having a higher level of expected happiness from childbirth, which is a proxy for a positive family imaginary, not only helped respondents to remain faithful to their plans, it even encouraged them to increase their fertility intentions, COVID-induced uncertainty notwithstanding. Even if Italy, like other Western countries (Sobotka et al., 2021), is already showing the first signs of fertility decline, our results suggest that a homogenous reaction in terms of a downward revision of childbearing plans is unlikely as well. In a situation in which rational decision-making was rendered difficult by mounting uncertainty about the future state of the world, those individuals who strongly valued parenthood may have reconsidered their life goals in favor of plans to form a family.

The claim that narratives of the future play a causal role has been reinforced by our experimental analysis, through which we were able to assess the causal impact of a shared narrative of the future on fertility intentions. Indeed, respondents were randomly exposed to different future scenarios regarding the expected length of the pandemic, and we found their pre-treatment fertility intentions decreased monotonically as the expected length of the pandemic increased. Both experimental and survey data converge in supporting the hypothesis that the shadow of the future had a clear impact on shaping fertility intentions during the pandemic emergency.

Of course, our work is not without limitations. First, the retrospective analysis of intentions may be influenced by the reduction of cognitive dissonance between current and pre-pandemic fertility intentions. While the use of panel data would, of course, have been preferable, the exceptional and unexpected situation of the pandemic did not allow us to plan pre- and post-pandemic waves of a panel survey. Second, even if the Narrative Framework is built on top of the established pragmatist tradition, which has recently been reappraised by Beckert and colleagues, its operationalization within a survey is a novelty. While there are well-known and validated indicators available in the literature for “traditional” variables related to past experiences and psychological predispositions, status, and perceptions; for expectations, and especially for imaginaries, there are just a few examples of their operationalization (Vignoli et al., 2020b). For imaginaries, we relied on the literature on the “expected happiness” (Billari, 2009); however, the common finding that couples are already happier before the birth of a child (see, e.g. Myrskylä and Margolis, 2014) may be attributable to more than the perception of an increase in happiness from the arrival of a future child – e.g. it may be the result of having a happy relationship and high life satisfaction, which are positively associated with fertility (Mencarini et al., 2018). A more in-depth exploration of personal imaginaries of parenthood, together with their relationship to childbearing desires (Mynarska and Rytel, 2018, 2020), should be developed in future research. In addition, future studies should attempt to capture more directly the role of personal narratives of the future and their functions in the fertility decision-making process through the use of both quantitative and qualitative methods.

In this paper, we showed that when studying changes in fertility intentions after the onset of a pandemic, it is important to take shadow-of-the-future factors into account. However, our conclusion that the shadow of the past is substantially less relevant than the shadow of the future for understanding the adaptation of the fertility decision-making process in a situation of increasing uncertainty may be traced back, at least in part, to the fact that our empirical analyses focused on the short-term effects of an unprecedented uncertainty shock. It is possible that long-lasting economic hardship due to subsequent waves of the pandemic and the responses to it may become increasingly important, and thus influence potential reductions in fertility. In addition, the shadow of the past already played an important role in shaping pre-pandemic fertility intentions. However, we argue that our results are suggestive of the factors that drive fertility decision-making in contemporary societies, even beyond the pandemic emergency. Long-term societal changes driven by globalization, and the “harsh new world of economic insecurity” (Hacker, 2019, p. xvi) that have accompanied them, are likely to have made narratives of the future more salient for fertility decisions in contemporary Western societies, in part due to the increasing pervasiveness of the media coverage of the economy. The Great Recession, for instance, fueled general perceptions of uncertainty about future economic conditions that may have hampered fertility even in countries and social groups that were only marginally affected by mass lay-offs or company bankruptcies (Sobotka et al., 2021; Hofmann et al., 2017). Preliminary empirical evidence corroborating this argument has been provided for the US, where it has been shown that fertility rates at the state level have been influenced by unemployment rates at the national level, and the press coverage of the economy, net of state-level economic conditions (Schneider, 2015). Future research should clarify whether the spreading of narratives of an uncertain future is responsible for the homogeneous decline in fertility that the US and European countries have been facing since 2010.

In the next screen we will provide you with up-to-date forecasts concerning the evolution of the coronavirus pandemic.

Within the last few days there haven't been substantial variations in the number of infections, hospitalizations and deaths. The task force composed of leading experts on the coronavirus pandemic eventually obtained reliable predictions about the future of the pandemic in Italy.

The experts predict that the coronavirus pandemic emergency will last X before a return to normality.

(5 randomized scenarios for X: 3 months, 6 months, 12 months, 2 years, more than 2 years.)