Young individuals as microcosms of the Portuguese crisis

ABSTRACT The Portuguese crisis affected the country’s collective identity, and ‘the timing of life’ at which it struck individual lives in this case is also significant. Quantitative figures show that young people were particularly affected by this crisis. However, a long-run qualitative approach provides a multilayered and quite complex view of what this crisis is embedding in young people’s lives and minds. In qualitative research on ‘middle class’ transitions to adulthood carried out in 2009, 52 young adults were interviewed about their educational, residential, occupational and romantic lives. In a follow-up study, these individuals’ trajectories, plans and expectations are now updated; their past and present confronted; and effects of the crisis on their lives questioned. The discussion is held in the form of a critical approach to the theories of individualisation, and goes to the heart of the ‘generation in itself’ vs. ‘generation for itself’ and ‘biographies of choice’ vs. ‘discourses of choice’ debates.

dynamics, and this crisis was far from an exception. That young people and young adults were thus particularly and negatively affected by the crisis, and more specifically by the austerity measures taken in its name in the past few years, is as far as static quantitative figures have been able to go in informing us about unemployment rates ( Figure 1, as an example), the increase in the young NEETS category, and the duration and persistence of precarious ways of living and working. However, long-run and qualitative accounts of the same reality give a both more accurate and more complex view of what this crisis may have been embedding in young people's lives, minds, ideals and dreams (Nico, forthcoming).
Researchers carrying out qualitative and biographical research in 2009 on therather broadly defined -'middle class' transitions to adulthood were somewhat surprised by these historical circumstances. Fifty-two young adults aged between 26 and 32 were interviewed about their educational, residential, occupational and romantic lives in an attempt to map and typify transitions to adulthood as socially stratified phenomena (Nico, 2011). The crisis itself did not play the role of a protagonist in the interviews and trajectories at that point. As a concept, in 2009 the crisis had only taken on an abstract outline. Its effects were not yet felt or imagined, but merely speculated on and not concretely experienced. This serendipity effect of the crisis, the Troika intervention and the measures taken by the government in response, inspired a follow-up study which involved mobilising a multiplicity of life course research instruments (biographical interview, life calendar, focus of control exercise, past reality checks) to re-interview the 2009 participants. This took place in 2016seven politically dense years after the first interviews, and in a new political context. New compared to the previous one, in which the government had been led by the centre-right PSD (Social-Democratic Party), which many would characterise as conservative in values and neo-liberal in its views of the economy and welfare; and also compared to the last three decades, in that the 2016 scenario represents the first time since the democracy-seeking revolution of 25 April 1974 that the country's centre-left parties have agreed to agree on important social issues related to inequalities, poverty and the population's well-being.
Focusing mainly on the second part of this ongoing study, this article will begin by presenting the research design, methodology and data used and suggesting that the eclectic and macro nature of life course methods advocate them as a privileged way of collecting Figure 1. Unemployment rate among 15-24 year olds, by gender, in Portugal andUE28 (2002-2015) Source: EU Labour Force Survey (in Carmo and Matias, 2016). qualitative data on real people and their real lives. I subscribe to the view of Thomas and Znaniecki when they say that the superiority of life records over every other kind of material for the purpose of sociological analysis appears with particular force when we pass from the characterisation of single data to the determination of facts for there is no safer and more efficient way of finding, among the innumerable antecedents of a social happening, the real causes of this happening than to analyze the past of the individuals through whose agency this happening occurred. (Thomas & Znaniecki, 1918, pp. 294-295) This is achieved by attempting to recover the 'historicality of the individual' (Abbott, 2001) and by looking at individuals as 'microcosms' of societal phenomena (Bertaux & Delcroix, 2000) or historical landmarks. Individuals are the walking crisis; they are its crystallisation on the time-space axis. This attempt is valid for both the first and second parts of the research, but goes in the opposite direction to that taken in the post-modern approach to narratives, where 'any intention of interpreting society "as it really is"' is abandoned, 'shifting to post-modern or narrative approaches, in which the interview text replaces society as the focus of study' (Thompson, 2004, p. 238).
At the intersection of youth with the importance of temporality in the research design, and within the framework of our research, there are three studies of note. One of these was an absolute pioneer in longitudinal research, which combined historical time with individual time, analysing the social trajectories of individuals from their childhood during the Great Depression until they reached adolescence, 'tracing step by step the ways in which deprivation left its mark on relationships and careers, life styles and personalities'. In this work by Elder (1994) we find the theoretical and methodological sustenance for the study of individual trajectories as bearers of national history. This is also what we are seeking to do in the present study (albeit with an interval which is much shorter, but which Rindfuss (1991) says is demographically much denser: that of the transition to adult life). The other two projects were also conducted on a large scale, attracted great attention and are again Anglo-Saxon. Despite being longitudinal, they are more contemporary, both theoretically and methodologically, differing from that of Elder mainly with regard to their theoretical stance. While Williamson's (2004) work, which revisits a group of troubled youths in a Welsh community 20 years on, is admittedly 'grounded theory' and a 'personalised' investigation as a consequence of the relationship the researcher developed and maintained with the participants, the study 'Inventing Adulthoods' took place over a 10-year period within a territorially dispersed project undertaken by a large and dedicated team, and followed strong theoretical direction in both its hypotheses and its analysis (Henderson, Holland, McGrellis, Harper, & Thomson, 2007/ 2009Thomson, 2009;Thomson et al., 2002).
Following this presentation of the details of the research design and instruments and the arguments behind them, two sets of results will be discussed. One has to do with the most relevant changes and continuities in the values and practices of these individuals across the past seven years, asking how they identify and perceive their 'strategic adaptation' (Elder & Giele, 2009, p. 14;Giele &, Elder 1998, pp. 9-10) to economic and political circumstances. To this end we used the whole life calendar, biographical interviews and especially the innovative 'past reality check' data collection and production instrument. The aim was to analyse the process of subjective and objective life course change; the main result showed the very small extent to which work ethics and life values have been affected, in the sense of corrupted, by the unfavourable economic scenario. The second has to do with the interviewees' recent life story and how it relates with the crisis, and the extent to which individuals saw their life events as satellites of the contextual crisis. This used the updated life calendar, the biographical interview and the innovative 'locus control' exercise instrument, in which interviewees are able to identify key moments in life and correlate them to some extent with the crisis. Counter-intuitive results were found in relation to the ahistorical understandings of their biography.
The conclusion will sum up these results, calling attention to how qualitative understandings of highly quantifiable phenomenasuch as financial or economic crisesare necessary if one is to grasp the nuances of a historically situated biography, which vary and oscillate from the identity to the explanatory natures of narratives, or in other words, between 'realist' and 'neo-positivist' approaches to life stories (Miller, 2000, p. 13).

Methodology and data
Albeit using new data gathered by different data collection and co-production instruments, the 2016 follow-up project is a natural continuity of the epistemological premises considered in the 2009 research. The 2016 methodological design follows the same epistemological statement, as it goes to the heart of the qualitative and longitudinal concerns and debates about the collection, interpretation and validity of data at specific historical moments. Its instruments are deeply grounded in the four well-established principles of life course researchagency, linked lives, timing of lives and cultural and historical location (Elder, 1974(Elder, , 1985Elder, Johnson, & Crosnoe, 2002;Giele et al., 1998); it combines elements of qualitative and biographical approaches with quantitative-oriented instruments of inquiry and the co-construction of life stories, such as the life calendar (Nico, 2016a). Most of all, it advocates the need to resort to qualitative data in order to fully understand how much of the crisis is embedded in young peoples' lives and minds, and to longitudinal data capable of confronting and comparing past and current interpretations of the courses of lives. Only by doing so can sociological flesh be put onto the statistical bones produced by statistics institutes and used and abused by the media and political discourse.
The earlier part of this research followed the premise that transitions to adulthood are a moving target, especially vulnerable to rapid social change and severe economic oscillations; and also that as an object of study, such transitions are particularly vulnerable to the biographical turns in the social sciences, as well as to the acritical affiliation with certain individualisation theories (namely Beck, 1992) to which some authors have been calling attention (Roberts, 2012). Trying to avoid some of the 'fetishism of the present' linked to this vulnerability, in which 'youth as a concept prompts researchers to often focus only on the "here and now" rather than taking a longer-term process view' (Goodwin & O'Connor, 2015, p. 39), the research approach was more easily integrated into the life course perspective than into youth studies as such, highlighting and using the concept of generation more than that of age group. The latter thus became more of an operationalisation tool than a thematic in itself.
One important argument underlying the research was based on a critical approach to the theory of individualisation. This is frequently used as a 'package', in which all contemporary changes in a life course, such as deinstitutionalisation, destandardisation, pluralisation and differentiation, are understood as moving in the same direction or at the same pace, independently of the social and geographical coordinates at which the individuals are situated (Brückner & Mayer, 2005;Nico, 2015). As such, in the 2009 and 2016 projects, as far as possible the different dimensions of the life course were analysed through different lenses, avoiding any self-contamination and reciprocal bias of the results (Nico, 2014;Nico & Caetano, 2017).
In the first part of the project, whose fieldwork was conducted over the course of 2009before the effects of the crisis (and the ensuing austerity measures) were concrete and visible in people's lives -52 biographical interviews combined with life calendars were applied to 'missing middle' class young people in Portugal (Roberts, 2011). This set of interviewees encompassed a significant, but not statistically representative, variety of individuals in terms of educational attainment, residential status and family social background. The sample was also gender-balanced. The major concern was to identify and map different trajectories towards adulthood, along with the social backgrounds and contexts that shaped and determined them (Nico & Caetano, 2017;Nico & de Almeida Alves, 2017).
The second, still ongoing, part of the study is recovering the contacts made in 2009. The approach has been direct and the agreement to re-participate total. No attrition effect has so far been detectable, and 13 adults, now aged 32-38 years, have been already interviewed. All these interviewees have been pleased to be re-contacted, consider the exercise of comparing past with present very entertaining and useful, and have demonstrated a relative eagerness to share how their life has turned out and a relative curiosity to know how the researcher's life has fared seven years on. Overall, ours is an intimate and informal conversation with a 'known' stranger. This has again allowed the researcher to be perceived as a 'particular kind of confidant, the kind that disappears after the interview' (at least for seven or more years), and to whom secrets can more easily be told, making him/her the 'receptor of words to which not even those the closest to us have access to' (Lahire, 2002(Lahire, /2004).
The following table shows the range of life course instruments employed in the 2009 and 2016 sets of research (Table 1). Both used the combination of the life grid and the biographical interview to improve qualitative and subjective data collection and analysis, while simultaneously seeking to maintain the actual order of the narrated events and thereby escape Bourdieu's 'biographical illusion' (Bourdieu, 1994(Bourdieu, /1997Nico, 2016a). In this regard, Bourdieu states that the narrative, whether biographical or autobiographical, for example, the discourse of the interviewee who 'opens up' to an interviewer, offers events which may not all or always unfold in their strict chronological succession (anybody who has ever collected life histories knows that informants constantly lose the thread of strict chronological order) but which nevertheless tend or pretend to get organised into sequences linked to each other on the basis of intelligible relationships. The subject and the object of the biography (the interviewer and the interviewee) have in a sense the same interest in accepting the postulate of the meaning of narrated existence (and, implicitly, of all existence). (Bourdieu, 1994(Bourdieu, /1997 The interference of the life calendar in the interview, as an instrument that is capable of challenging previous and pre-determined views of the past, and stimulating new reflexivity on the causal and emotional links between events of lifelinks that might be 'discovered' during and because of the interviewis a positive one. It allows the biographicalobjective and subjectiveinformation to be shared in a more detailed and accurate format. At least since 1999, more research and publications on the combination of the life grid with semi-directed interviews have been presented as leading to new forms of understanding and qualitative results (Parry, Thomson, & Fowkes, 1999). Despite this, and although the life grid is very often employed in life course research, its combination with qualitative and biographical interviews is still not that common. Previous uses do show, however, how this combination enables the relationship between the individual experience and the identification of historical or individual key moments to improve the transparency of the relationship between the interviewee and the interviewer/ researcher, and allows a holistic understanding of the phenomenon by promoting the interdependence of the rapports about the various events and facilitating the recapitulation of the life course narrative (Parry et al., 1999).
In an attempt to make the most of the new, updated and re-reflected material on the 2009-2016 window of observation, the 2016 research used two new data collection and co-production methods: the 'past reality check' and the 'locus control exercise'.
The instrument I have called 'past reality check' allows for an adjustment between the 2009 and 2016 narratives (about values and practices). It consists of a selection of polemic,  representative and/or paradigmatic quotes from the 2009 interview, which in some way demonstrate the most representative issues, opinions and/or plans at the time. The goal of using these old quotes is to act as an icebreaker and to set the past as the reference for the discourse, making it harder to completely re-write the (recent) past. As developed below, the quotations selected so far mostly had to do with: work and consumption ethics, homeownership and renting, and emotional rupture cycles.
The instrument I have called 'locus of control exercise' is partly inspired in psychology. It allows the individual to situate her/his most relevant life events on a scale from agency to structure (Table 2). An effort was made to make this scale correspond to the life course principles, such as agency, linked lives, and cultural and historical location, and these principles were adapted to the research design.
The scale is conceived as starting with a category of agency, where individuals are seen as completely responsible for their own actions and options, 1 followed by the linked lives category, which assumes that each individual's actions have intra-and inter-generational consequences. 2 However, in other cases this correspondence between the operative scale and the life course principle is not that linear. This is because one of the principles of life course research is broad enough to merge social, economic, geographical and historical backgroundsthat is, to sum up a substantial part of the sociological-analytical framework of social differentiation. As Elder and Giele put it, 'this principle underscores the multiple layers of human experience, the social hierarchies, cultural and spatial variations, and the social/biological attributes ' (2009, p. 12). In the same text, by giving an example of this life course principle, these important life course authors distinguish between the level of 'the Great depression', shared by all (which would correspond to the Portuguese Crisis' in the present article), and othermore socially groundedvariables, which differentiate the effect of the historical context of the Great Depression. Similarly, in the present article the latter variables would correspond to the interviewees' 'socio-economic situation'.
Under this life course principle of the 'historical time and place', the remaining correspondence used in the data collection instrument included a social location category that considered close social context, 3 followed by a historical location category in which the crisis was the only element.
Lastly, and outside the scope of the life course research, one category considered that the 'unobserved heterogeneity' in statistical analysis simply corresponded to 'chance'. 4 The challenge put to the participants was to quantify in percentage terms the effect of each of these categories on their relevant life events (Table 3). Immediately before their interviews, interviewees were asked to provide a list of 'relevant' life events.

Emergent concerns and changes: lives and mindset
This set of results is mainly based on the 'past reality check' instrument, with three emergent trends found. The first has to do with the levels and values associated with consumption, use of credit and loans and savings practices or ambitions. Many of the 2009 interviewees demonstrated a very strong aversion to debt and would not have or use credit cards, even if it might have helped them overcome some difficulties towards the end of each month. Using credit cards or taking out loans was not considered an option, even when bank accounts were dangerously close to zero. In 2016, however, although these interviewees' discourse was very similar and true to the core values transmitted in 2009, practices had been slightly adjusted. This adjustment is socially stratified. On the one hand, for those with less qualified 5 and badly paid 6 occupations, credit card use had become more flexible. Although not on a daily basis, credit cards were now used to schedule holidays or travel, the expectation being to pay off the debt immediately on receipt of the extra holiday or Christmas 'month' (by law, salaried Portuguese workers receive 14 months of pay, with the extra 2 paid before annual holidays and before Christmas), and with the reassurance of being in a common economic partnership (civil union or marriage). In these cases, this was done to maintain a certain 'young' lifestyle that is more common in couples (as yet) without children. On the other hand, among individuals with conservative financial values who had managed to achieve an ascendant social mobility trajectory in the past seven years and enter a consensual union, the adjustment to the crisis was surprisingly higher. Although possessing the means to maintain or improve a certain lifestyle, these individuals tended to be hyper-conscious of the crisis, reducing their consumption habits to a minimum and developing saving strategies involving low-rate, low-risk investments or early payment of mortgages.
Summing up, for those with the same 2009 values regarding consumption, savings strategies and resort to credit, (i) the individuals with more precarious and/or worst paid occupations adjusted their values to practices oriented towards their daily lives, aiming at improving them and from time to time having a holiday or other 'luxury' that was normally out of their financial league; (ii) while the ones with more stable and/or well-paid occupations tended to reorient their consumption values and practices towards the future by saving money and/or anticipating mortgage payments.
A second emergent trend is linked to a dividing issue in 2009: opinions on homeownership and renting. In 2009, some interviewees were strongly against homeownership; all those who have already been re-interviewed in 2016 had changed both their views and their practices in this respect. From a significant original aversion to buying a home, which was said to create a life-long debt, which they would always live to pay off and would be an inhibiting factor for a putative geographical mobility or migration project, these individuals were now more tempted to become homeowners (some already had). The three motivations for buying a home are well documented in the literature and apply, autonomously or conjointly, to these cases (Winstanley, Thorns, & Perkings, 2002, p. 815). One set of motivations has to do with the lifecycle model in which a change of residence serves the need for more space to accommodate the changing age and sexual composition of the family structure. The second set of motivations has to do with economic rationality, in which these residential changes are treated as the result of economically rational decision-making. Finally, a third set of motivations to change or buy a home is the environmental aspect, meaning that people choose to buy or change their residence according to where and near whom they want to live. All this means that the lifecycle effect has maintained a certain predominance, even in contexts of crisis, leading these individuals to buy apartments near their families of origin and with the aim of accommodating their own families in the future.
These sets of motivations are not mutually excluded by the participants, but instead appear conjointly in their discourses, in a process of accumulating reasons that together increase their propensity to see homeownership as positive and act accordingly. The profile of this set of re-interviewees may also help explain this tendency towards homeownership. They included people who had already left home in 2009, in order to live with a partner or spouse, live alone or live with friends or flatmates, so they had already taken the plunge of leaving the parental home and committing to a monthly obligation to pay rent. What they rejected was the long-term commitment of a mortgage, which they refused because of the uncertainty they perceived in their lives at the time, the relative flexibility they wanted to have in their lives, and also in some cases, because of left-wing anti-property political views (this was the least important reason). So while this set of reasons explains why these interviewees opted for rented apartments when they decided to leave their parental home in or prior to 2009, the combination of the lifecycle model, the economic rationale and the environmental aspect explains why they had experienced an increased tendency to choose homeownership seven years later.
The third emergent trend entailed voluntary and involuntary rupturesmainly involving professional well-being or romantic relationships. In 2009, some participants displayed a low tolerance for precarious situations, be they emotional of professional. They considered rupture to be a positive action, a necessary harm, a long-term investment in happiness. In 2016, these and other participants displayed this same approach to life, but with some distinctions with regard to the particular sphere of life in question. While in 2009 rupture cycles were more directed at emotional and romantic trajectories (Nico, 2016b), in 2016 they mainly concerned the professional sphere. In fact, especially for those who began working very young, before or during their university studies, longterm precariousness had become intolerable. They had proven to themselves and others that they are working-class individuals, with early achieved independence and residential autonomy, but they had also gathered information about their rights as workers, their entitlement to well-being, and the processes that can lead to burnout. They now had 'chosen' to say 'enough is enough' to this long-term precariousness, and as a consequence had experienced transitional unemployment episodes. A couple of quotes will illustrate these intolerable experiences of precariousness and moments of rupture and non-conformity.
Yesterday I heard a guy on the news saying 'it's better to have a precarious job than to be unemployed'. Hmmm, really? Not in my opinion. If you think you can be precarious for 9 years, like me. When I started working there I would cry because of the physical tiredness, I was not used to being on my feet the whole day, I only had one day off a week … But then the mental tiredness came. The work environment is bad, people are unhappy working there and I would sometimes get home in a really bad mood (for my daughter). And I thought; I don't want this for me, I need to get out, need to leave this environment. And you know? It is a relief, to have the possibility to start something new. (Julia, bachelor's degree, unemployed from fashion shop, rented apartment, civil union, one child) There are many forms of non-dignifying work, and people have to know about that. I and everyone else from Architecture knew that. It was fantastic: we had to have our own car, computer, two languages, Autocad, of course. 'Come here, I'll let you work here!'. I found that statement about being "better to have a precarious job than to be unemployed. I thought 'yes, it's also better to have hepatitis B than aids, but the best is to have none of that, right?!'. (Jorge, almost graduate, bank manager, single, homeowner)

Serendipity: contractions and red flags
Some counter-intuitive results emerge from the analysis of the locus of control exercise, especially bearing in mind that it came at the very end of each interview, in an attempt to summarise information shared in the previous couple of hours. The information analysed here, although regarding the same 13 re-interviewees, refers to 66 relevant life events listed by them. After each interview, these were coded as to their impact (positive/negative/ambivalent), their dimension of life (family, work, school, health, home, other) and their occurrence (relevant events/relevant omissions of events 7 ).
The list of relevant life events in the past seven years is the first source of surprise. Transition to adulthood is indeed a demographically dense period of life (Rindfuss, 1991), but some spheres of life are denser (more events per period of time) than others. There is a delay in entering conjugality in Portugal, a high age at the first child, and a very low birth rate, while precariousness and informal work are important characteristics of the Portuguese labour market. Additionally, these aspects were emphasised during the crisis period. As such, and particularly given the latter aspect of the situation, one might have expected most relevant life events to be work-related. However, the data reveal a big gap between the number of work-related events in the interviewees' lives and the number they considered relevant. Almost half the life events deemed relevant were family-related: romantic relationships (beginnings or breakups), parenthood, marriage, divorce and so on, whereas only about a quarter were work-related. To an extent, the 'persistence of the family life cycle' (Elchardus & Smits, 2006) is thus resistant to the economic crisis. In a way, it is the events that have the least to do with the crisis that are considered relevant, are part of the biography of identity that these individuals choose to embrace and share. In their discourses, there is a denial of the crisis as central to their identity. The crisis does not define them.
When it comes to the effect of the crisis, you can write down 'zero' for everything. The crisis did not have any influence on my decisions and career. I even think I grew in those periods of alleged crisis. (Ricardo, graduate, car salesman, homeowner, married, one child) The financial crisis does not have any effect in my career. If it did, it would have ruined it. (Jorge, almost graduate, bank manager, single, homeowner) Family events were mentioned more than any other as having been important in our seven-year window of observation. This may be linked to two interrelated aspects. One is that overall, the participants tended to list positive aspects and events more than negative ones. When summarising their seven-year life period, they tended to attach a positive view to their biography, mentioning the aspects that, 'at the end of the day', mattered most to them and best defined their progress in life. This is an act of implicit resilience which, when it comes down to it, supports their discourse and their interpretation of life. The second aspect is related to the first: the most frequently mentioned family events were in fact turning points and/or irreversible events in people's lives. By this, we are (and the participants were) referring to events such as marriage, and having a first or second child. 8 A second source of counter-intuitive results has to do with the fact that between school/studies, family and work, the latter is actually the one that returned the lowest proportion of negative impacts. Unemployment was the event with the greatest responsibility for this result, in two ways: by not even being mentioned as relevant (compared to other life events); and by being mentioned as a 'positive' event (in the long term). The shared stories demonstrate how these unemployment events can be seen as positive. The reason is again related to long-term precarious situations in the labour market, which after several years reach the point at which they are unbearable and 'force' the individual to take a pause on the road to burnout or depression. This ability to 'pause' enabled interviewees to subsequently see this as a positive action, which was necessary in order to break dangerous cycles of long-term precariousness, decreases in emotional wellbeing, conformism and identity marasmus.
A third set of unexpected results can be seen from Table 4, which shows the distribution of the mean percentage attributed to the effect of the different categories of the agencystructure scale on relevant life events. We can see that with regard to general issues, the impact tends to decrease on the agency-structure scale. This means that generally speaking, the individuals believed it was their own agency (48.6%), followed by their families and friends (26.7%), their social context (11.2%) and only then the crisis (7.1%) that determined their life events. 'Chance' is the only predictor that scored below the crisis. This might mean that predictors for meaningful life events which fell well outside the individuals' control may have been left out of the narrative accounts about their trajectories, regardless of the actual impact they actually did or did not have in terms of shaping their life course.
Two other aspects of note are concentrated in the 'the crisis' column. In fact, according to the individuals' views, of all the influences their professional trajectories may have been subjected to, the crisis was the least important (3.7%). In their discourses, interviewees tended to individualise the explanation of their trajectories: their actions, their bosses' decisions, their co-workers' attitudes. Even the negative work-related events can in a sense be read asocially, as if the crisis provided not a context, but an interpretative vacuum. On the other hand, the effect of the crisis on 'other' events is significantly high (18.9%). The crisis belongs to 'others'. Unemployment of parents, bankruptcy of a relative, migration of a best friend 9 the examples are numerous. Interviewees displayed a somewhat bipolar approach to the crisis: they were bystanders with regard to the effects of the crisis on the 'other' people in their lives, and they were (or became) immune to the crisis themselves. It is nonetheless important to highlight that in this case the effect of linked lives is considered negative; the crises of others, in which family members occupy an important quantitative and qualitative place, are indeed those that illustrate the effect of the crisis in more detail.

Conclusions
This follow-up study means that the trajectories, plans and expectations of young Portuguese adults have not only been updated, but past and present were literally confronted with one another in a co-production interview. At a certain point, the individuals were asked to specifically reflect on and quantify the effects of the crisis on the several dimensions of their lives in the previous seven years. Results go to the heart of the discussion on a distinction between 'biographies of choice' and 'discourses of choice' (Nico & Caetano, 2017). This is because in 2009 the crisis was slightly present in discourses, but absent from actual trajectories, but in 2016 the contrary was verified. The presence of the crisis and its many effects at the micro-individual level was clear in the 2016 interviews, in which the events and conditions in the labour market were identified, described and sequenced; but it is also evident that this protagonism disappeared when the interviewees were challenged to interpret their life events as satellites of the crisis. This apparently asocial reading of their life trajectories, which is not the same thing as a lack of political literacy or political participation, reinforces the concern about the 'epistemological fallacy of the late modernity' as Furlong and Cartmel eloquently put it when they said that 'blind to the existence of powerful chains of interdependency, young people frequently attempt to resolve collective problems through individual action and hold themselves responsible for their inevitable failure' (Furlong andCartmel (1997/2007, p. 144).
There is also evidence that this fallacy may possess a politically significant meaning, in the sense that it could be a result of neo-liberal discourses which uphold the idea that 'we are what we choose to be' 10an idea that helps produce a generational discourse which denies the context in an individual's own biography, thus neglecting its historical location and interpretation. Although probably involuntarily on the part of some influential sociologists, individualisation theories fit the neo-liberal discourse like a glove. After all, (young) people would like to acknowledge themselves as authors of their own lives, and the canvases on which they visualise and project their life goals, ambitions and plans are equally white for everyone, be they middle or upper class, rural or urban, poorly or highly educated, etc. What this research shows is that the appropriation of individualisation discourses by political agents may not be a one-way street, but part of a circular movement in which the dissemination of this individualised neo-liberal discourse is able to re-contaminate the individual's discourses and ways of thinking.
These conclusions would not have been possible without a wide window of observation and a combination of quantitative and qualitative approaches to life stories. 'Time matters' (Abbott, 2001). Phenomena such as long-term precariousness in the labour market are invisible in the statistics collected about young people; and phenomena like this are responsible for the apparent contradictions in young people's discourses and biographies and an apparent lack of agency and action. People are indeed capable of providing an explanation for their actions, and in that should lay our concerns as social scientists and citizens. 'If men define a situation as real, it is real in its consequences' (Thomas theorem).