Poverty as a Driver of Stigma among Finnish Children during the Covid-19 Pandemic– Evidence from the 2021 Children’s Voice Survey

Stigma is a social problem that can have severe outcomes on both individuals and society. Previous research shows that children’s experiences of stigma may increase during times of stress and hardship such as the Covid-19 pandemic, but also that poverty and a lack of resources can create stigmatization. However, both stigma and poverty are multidimensional phenomena that have been investigated mainly among adults using single indicators. This article contributes to literature by studying children’s experiences of stigma and its relation to poverty by using a multidimensional approach. More specifically, it investigates Finnish children’s experienced and internalized stigma during the second year of the Covid-19 pandemic, and analyses how these dimensions are associated with subjective poverty and material deprivation. We use data from the 2021 Children’s Voice survey conducted by Save the Children, Finland. The results show that both dimensions of stigma are correlated with subjective poverty as well as material deprivation, even when controlling for socio-economic and other variables. Also low psychological wellbeing is a significant driver of both forms of stigma, while living in a one-parent household was significantly associated with internalized stigma, but not experienced stigma. By contrast, while higher self-esteem seem to reduce both forms of stigma, a higher number of good friends was found to only reduce experienced stigma. As there are both short- and long-term negative outcomes of stigma, for example in terms of mental health problems and social exclusion, child poverty should be taken seriously – especially during times of crises such as pandemics.


Introduction
Finland is a part of the Nordic welfare model, known for its generous and universal social policy and its low degree of poverty (e.g.Hakovirta & Nygård, 2021).It has also been ranked as one of the happiest countries in the world several times (Helliwell et al., 2023).However, as in many other countries, the Covid-19 pandemic brought along several challenges to this model, but also to ordinary people and their children.For instance, in the first phase of the pandemic, economic activity stagnated and many people lost their employment.Statistics from Eurostat. (2023a) show that during the Covid-19 pandemic's first year (2020), the unemployment rate in Finland climbed from 6.8% to 7.7%.As for children, the closing of schools and cancellation of hobby activities for schoolchildren increased a social vacuum that increased both loneliness and stress (Pekkarinen & Miettinen, 2021;Varanka et al., 2021).Moreover, the pandemic also had economic effects on children.For instance, the percentage of children living in poor households increased from 10.3% in 2019 to 11.6% in 2020 (Eurostat, 2023b).Socio-epidemiological research also shows that children became increasingly faced with lower wellbeing and health, weight problems, increasing mental health problems and lower levels of self-esteem during the first phase of the pandemic (e.g.Adibelli & Siimen, 2020;Cusinato et al., 2020;Fiorillo & Gorwood, 2020;Kauhanen et al., 2022).However, what still remains understudied is the prevalence of experienced stigma among children in a pandemic context, and to what extent such experiences are related to poverty and material deprivation.
Stigma refers to the discrepancy between social and personal identity (Blaine, 2000), and stigmatization is the process of social re-categorization or depreciation that renders a person or a group a disadvantage position in relation to others (cf.Reutter et al., 2009;Sutton et al., 2014;Walker, 2014).Previous research shows that stigmatization is closely related to experiences of poverty among the adult population, for example in terms of social depreciation or exclusion due to inferior consumptive resources (e.g.Reutter et al., 2009;Sutton et al., 2014).Some evidence suggests that this is also the case with children (e.g.Knifton & Inglis, 2020;Baiocco et al., 2019;Ridge, 2002Ridge, , 2011)).However, as the bulk of research on stigma has been conducted among the adult population, we do not know very much about the prevalence and expressions of stigma among children, or to what extent stigmatization is driven by poverty.
This article seeks to fill this void by investigating the prevalence of two dimensions of stigma, experienced and internalized stigma, among Finnish 12-17-yearolds during the second year of the Covid-19 pandemic, and how such experiences are related to two different aspects of poverty: subjective poverty and material deprivation.According to Mickelson and Williams (2008), experienced stigma relates to perceptions of being stigmatized by others through, while internalized stigma has to do with feelings of self-depreciation and negative self-images (Mickelson & Williams, 2008).The first poverty indicator refers to children's evaluations of their own family's economic resources (cf.Hakovirta & Kallio, 2015), whereas the latter relates to children's perceptions of their consumptive capacity in relation to others 1 3 Poverty as a Driver of Stigma among Finnish Children during… (Main & Bradshaw, 2012).We analyse data from the Children's Voice Survey conducted by Save the Children, Finland, in the spring of 2021 (N = 1102).This data was collected electronically among 12-17-year-olds in different regions of Finland and provides cross-sectional data on Finnish children during the pandemic.
This article makes at least three important contributions to the literature on stigma and its relation to poverty.First, since stigma is ordinarily something that has been studied mainly among adults, this article brings in the perspectives of children.It also uses a child-centred methodological approach that brings out children's own voices, not the voices of their parents or teachers.Second, it uses a multidimensional approach to the understanding of stigma, but also its relation to poverty, which is still quite rare in social and child indicator research.Third, since the bulk of what has been written on stigma and poverty refers to liberal welfare regimes with less public service coverage and higher poverty rates (cf.Ridge, 2002), the Finnish case, with its universal coverage of public services, free schooling system and low poverty rates, can provide important knowledge on the extent and drivers of stigma in a social-democratic welfare regime (Esping-Andersen, 1990).
The rest of the article is structured in the following way: we start with a discussion on the concepts of stigma and poverty, and what we know from previous research about the association between the two when children are concerned.On the basis of this, a number of hypotheses guiding the empirical study are posed (see below).In the following section, the data and methods are discussed, and in the penultimate and final sections, we present the findings and conclusions.

Stigma and Stigmatization
According to Goffman (1963), stigma is a deeply discrediting attribute, in which a subject is reduced and discounted, both socially and psychologically.Reutter et al. (2009), as well as Lister (2004), connect stigma to the wider ideological construction of poverty, according to which the poor are seen as burdens and undeserving of societal assistance.Hence, stigma can be defined as a discrediting social label or attribute that may have undermining effects on a person's or a group's self-image, identity and self-confidence -often with negative outcomes on other areas of one's life, whereas stigmatization is the process of social re-categorization or depreciation that renders a person or a group a disadvantage position in relation to others (cf.Reutter et al., 2009;Sutton et al., 2014;Walker, 2014).These processes can be triggered and orchestrated by either external factors, for example some kind of socially excluding action such as bullying, or by internal factors, for example feelings of shame or depression (ibid.).For example, a person can experience stigma because of some particular attribute, but becomes stigmatized when such feelings begin to have effects on one's self-image and wellbeing.In other words, a person not only becomes a part of the 'other' in society (cf.Lister, 2004), but also becomes internally and mentally burdened, since feelings of 'otherness' will harm one's selfesteem and make one question one's worth (MacDonald & Leary, 2012;Walker, 2014).
Correspondingly, previous research has identified at least two central, and interrelated, dimensions of stigma that captures this complexity (e.g.Mickelson & Williams, 2008;Reutter et al., 2009).These are the two dimensions of stigma explored in this study.The first dimension, which can be coined external stigma, or experienced stigma, relates to perceptions of being stigmatized by others through, for example, social exclusion and prejudice.In cases where there are a major discrepancy between the imagined or virtual identity, and one's actual identity, there is a risk for higher vulnerability and tension in interactions with others.The second dimension, which can be labelled internal stigma, or internalized stigma, is related to feelings and discrediting self-labelling, self-depreciation and negative self-images (Mickelson & Williams, 2008).
People deal with stigma in different ways.For instance, Moksnes and Espnes (2013) suggest that people with high self-esteem possess higher coping skills than others.Moreover, adults are likely to experience and cope with stigma differently than children.Generally, adults are more capable to handle stigma, while children have not yet developed a sense of self-esteem and are less capable to cope with stigma by their own (Orth & Robins, 2014;Ridge, 2002).As children get older, they become more aware of their family's situation, and can relate that to other's situation and to society in general (Lindberg et al., 2020).At the same time, their peers also become aware of such things, and it has been found that clothes (especially branded ones), and appearances are important for adolescent's social categorizations that might have a stigmatizing outcome (Fernqvist, 2013;Ridge, 2002).Geographical factors can also play a role for the prevalence of stigma.For instance, Ridge (2002) found that children living in rural areas were more exposed to stigma than those livening in more populated areas with higher degrees of anonymity (cf.Gubrium and Lødemel, 2015) .

Poverty among Children and young People
Poverty represents one of the oldest and most debated social problems of contemporary welfare states, and in a very general way, it can be said to refer to hardships and problems arising from an insufficiency of material recourses, that is, to be incapable to live to the life one wants to live due to lack of means (Alcock, 2006;Lister, 2004).Poverty has, of course, changed much over time, from being primarily a question of absolute poverty to being an issue of relative poverty, that is, to consider oneself as poor in relation to others (e.g.Townsend, 1979).Nevertheless, it is still a serious social problem, since it is related to important outcomes on health, wellbeing and other aspects of an individual's life, as well as on the overall societal structure and the degree of inequality (e.g.Alcock, 2006;Townsend, 1979).
Poverty among children and young people has been considered to be a more controversial problem than poverty among adults for two main reasons.Firstly, they cannot choose the family they are born into, or be expected to have control over their own incomes.Secondly, poverty is likely to affect children differently than adults, since they have less capacities and knowledge how to handle such problems practically and psychologically than their parents (Nygård et al., 2019; 1 3 Poverty as a Driver of Stigma among Finnish Children during… Chzhen et al., 2017;Gornick & Jäntti, 2012;Chen & Corak, 2008).Furthermore, and partly as a consequence of the abovementioned, experiences of poverty in childhood are likely to have long-terms effects, that is, to have adverse outcomes later in life, such as a higher risk of school drop-outs, anti-social or criminal behaviour in adulthood, or to become poor as an adult (Danziger & Waldfogel, 2000;Esping-Andersen et al., 2002).
However, poverty is a complex and a multi-dimensional phenomenon that can be defined and measured in several ways.One distinction that can be made is whether it should be seen merely as a monetary or resource problem, or as something else, for example as a problem of low participation or social exclusion (Schenck-Fontaine & Panico, 2019;Bárcena-Martín et al., 2017;Haughton & Khandker 2009;Laderchi et al., 2003).Yet the most common way of defining and measuring poverty in today's Western world is to use a monetary and relative method, that is, to assess how great a share of the population live in households with a equalized (harmonized) disposable income that falls under 60, 50, or 40 percent of the median for the whole population and for a given year (Nygård et al., 2019).As for children, this is not a very accurate measure, since it only captures children's 'theoretical' incomes based on their parents', or the household's, disposable incomes.Moreover, such poverty measures say very little about the depth or duration of poverty, let alone subjective aspects of experiencing poverty in some form (Hakovirta & Kallio, 2015;Harju & Thorød, 2010).Earlier research has predominantly used single, and single point in time measures of poverty or material deprivation, such as relative income-based poverty measures at a certain time and place, since this has been useful in comparative analyses for policy-making purposes.Still, this is problematic, since it does not necessarily capture the multi-dimensional character or structure of poverty, especially as children are concerned (e.g.Nygård et al., 2019;Schenck-Fontaine & Panico, 2019).
Accordingly, there a growing literature stressing the importance of measuring poverty multi-dimensionally, particularly in light of the UN Sustainable Development Goal 1, for example by using combining different measures of poverty, such as relative income measures or indexes on material deprivation (e.g.Nygård et al., 2019;Bárcena-Martín et al., 2017;Hakovirta & Kallio, 2015) or juxtaposing relative measures of child poverty to anchored ones (e.g.Chzhen et al., 2017).Moreover, the literature also recommends combining more objective indicators with more subjective ones, such as combining an assessment of subjective experiences of difficulties to make ends meet (e.g.Halik & Webley, 2011), with a child deprivation index that assesses whether children possess things they need or would like to own (Main & Bradshaw, 2012;Halik & Webley, 2011;Haughton & Khandker 2009).In this article we have chosen such a multi-dimensional approach, that is, to combine an assessment of children's perceptions of whether their families have enough money to make ends meet with the child deprivation index developed by Main and Bradshaw (2012).

The Relation Between Stigma and Poverty
The relation between children's wellbeing and poverty is well established in the literature (e.g.Gross-Manos & Bradshaw;Lindberg et al., 2020).For example, it has been shown that moving into poverty increases the risk of mental health problems for both mothers and children (e.g.Wickham et al., 2017), while Main and Bradshaw (2012) found that child deprivation is strongly linked to children's general wellbeing.The negative effects of poverty on children's wellbeing have been explained theoretically in different ways.One main explanation is that poverty and material scarcity creates stress within the family and this stress undermines child wellbeing (e.g.Conger & Elder, 1994).The same theoretical explanation can also be seen as relevant for children's stigma, although studies on children's stigma and its relation to poverty are rarer.There is some qualitative research on child poverty that has shown that poor children feel stigmatized (e.g.Ridge, 2002Ridge, , 2011;;Smith & Todd, 2019).Studies from Sweden and Norway (e.g.Fernqvist, 2013;Harju & Thorød, 2010), have also found that poor children are likely to experience stigma and shame.However, there are also factors having a modifying impact on this relation.One is social support or contacts: the more friends one has, the lower the risk of stigmatization (Boulton et al, 1999).Also children's attitudes towards consumption and the internalisation of materialist values can moderate the relation between poverty and stigma.Some previous research shows, for example, that poor children and young people generally tend to foster more materialist or pro-consumption attitudes than others (Chaplin et al., 2014), which in turn can increase feelings of stigma and shame.Strategies to avoid exposure to stigma can also mediate the relation.Walker (2014), for instance, found that one common coping strategy against poverty-related stigma was avoiding the 'normal', that is, to avoid places, social situations and interactions, which risk to reveal their poverty.Previous studies on child poverty and its social connections have also shown that one form of such avoidance among poor children is to avoid situations where their situation becomes more widely exposed, such as not to invite ones friends home or to avoid hobbies that are expensive (e.g.Easterbrook et al., 2014).
Based on the discussion above, a number of hypotheses can be postulated to guide the empirical analysis.First, following Mickelson and Williams (2008), we expect to find a higher prevalence of both experienced and internalized stigma in children experiencing subjective poverty or material deprivation (H1).Secondly, we expect to find that social and psychological factors, such as having friends or self-esteem reduces the link between stigma and poverty (H2).

3
Poverty as a Driver of Stigma among Finnish Children during… 5 Data and Methods

Materials
We used data from the 2021 Children's Voice survey that was conducted by the Save the Children, Finland (Pelastakaa Lapset, 2021).This survey was conducted as a nationwide online Webropol survey among Finnish 12-17-year-olds and could be completed between April 12 and May 2, 2021.The survey was made available to the children via various social media channels and was shared with a number of schools and education officials in Finland. 1 The data was collected both during school-time under teacher guidance, but it was also possible to complete the survey at other times or places.The survey was anonymous, which precludes any information about where, when and how each of the respondents completed the survey.The survey was available in both Finnish and Swedish and it contained a total of 49 questions (including demographic questions about age, gender, mother tongue, geographic region, etc.) focusing on issues of socio-economic situation, well-being, stigma, relationships, family conditions and other questions relating to life changes due to the Covid-19 pandemic.Most of the questions were either categorical questions (such as age) or Likert-scaled questions consisting of several items (such as stigma).
In total, 1102 respondents completed the survey, with the largest group consisting of 16-year olds (25%) and the smallest group consisting of 12-year olds (8%).The majority of the respondents were girls (76%), while 14% were boys and 10% of the respondents did not want to define their gender or answer the question.Almost all respondents spoke Finnish at home.The majority of the respondents lived in the south, in the mid-part or the northern part of the west coast of Finland (Pelastakaa Lapset, 2021).For the sake of our analysis, we excluded all respondents that had not answered all questions relating to our analysis, which ultimately left us with 730 respondents.
The sampling technique used in the Children's Voice survey could perhaps be described as a combination of a stratified, snowball and a convenience sample.Some children filled out the survey during class while others did it on their spare time.Due to the fact that the routines that were used during the collection processes varied, and since we cannot know how many children received the link to the survey, we cannot estimate a response rate, nor was it possible to obtain a strictly representative sample of Finnish children.For instance, some geographical regions, such as the Eastern regions, were not that well represented in the sample and the lion part of respondents were girls.Notwithstanding these issues, the sample can provide us with a fairly representative picture of how Finnish children experience stigma and poverty.

Data Analysis and Variables
The first phase of the analysis was to investigate the prevalence of stigma and its relation to poverty (H1).In the Children's Voice survey, stigma was measured with a 15-item scale asking children whether they, during the last six months, had experienced sentiments or emotions related to stigma due to their family's income situation and the response categories for each of these items ranged from (1) 'totally disagree' to (5) 'totally agree'.Following Mickelson and Williams (2008), we performed exploratory factor analysis (Principal Component Analysis with varimax rotation) of this variable to extract different underlying and more nuances dimensions of stigma.The factor analysis revealed three underlying dimensions.The first related mainly to stigmatization arising from one's relation to others, and the second to how children stigmatize themselves.By contrast, the third dimension related to positive emotions, such as senses of pride (see appendix 1), which is why we chose to omit the items displaying strong factor loadings on this dimension and to focus mainly on the first two dimensions.The next step was to construct two separate stigma variables on the basis of this.The first variable, experienced stigma, is a 4-item sum variable that reflects feelings relating to social depreciation and different treatment2 (Cronbach's Alpha = 0.900).The other variable, internalized stigma, is a 7-item sum variable reflecting feelings of shame and inferiority (Cronbach's Alpha = 0.875). 3he correlation (Pearson's correlation coefficient) between the two stigma variables was strong and positive (r xy .722;p < 0.001).In order to make a distinction between stigma and non-stigma, we imposed a cut-off line at 1.9 so that values in the range 1-1.99 represent non-stigma and 2-5 represent stigma.
Subjective poverty was measured by using the question "how easy/hard it is for your family to manage expenditures with the current income?" with the response categories ranging from 'very easy' (1) to 'very difficult' (5).To achieve a more balances response distribution we recoded these categories into three categories: 'easy', 'some difficulties' and 'difficulties'.Material deprivation, which refers to the lack of necessary components of an acceptable lifestyle for children (Main & Bradshaw, 2012) 4 was measured with 10 items relating to such consumption goods: new (not second hand) clothes, every-day availability of fruits and vegetables, having a smartphone similar to that of other kids in the same age, having an own computer, having an own room, having pocket money for one's own use, having a possibility to take 1 3 Poverty as a Driver of Stigma among Finnish Children during… part in regularly organized free time hobbies, having a possibility to go to movies, concerts or other events, having a possibility to organize a party in honour of special moments, and having a possibility to go on a vacation trip once a year.On the basis of these items, we calculated a 10-grade index of material deprivation according to previous research (Main & Bradshaw, 2012).According to Main's (2013) proposal of how to define the threshold for deprivation, we recoded this index into a dummy variable, where '0 or 1 missing item' constitutes non-deprivation, while 'two or more missing items' constitutes deprivation.
The second phase of the analysis was to assess the role of other factors for the two dimensions of stigma.Informed by previous similar studies (e.g.Schenck-Fontaine and Panico, 2019), we used two sets of independent variables: a set of social and psychological variables and the other representing background variables.Psychological wellbeing was assessed with the question: "How is your psychological wellbeing at the moment?" with the following response categories: 'good', 'quite good', 'average', 'quite poor' and 'poor' (Ryff, 1989). 5To receive a more balanced response distribution, we recoded the variable into a 4-grade variable, with two most positive categories combined.Self-esteem was assessed with the statement 'I like being the way I am' from Ryff's (1989) psychological wellbeing scale.The response categories ranged from 0 (don't agree at all) to 10 (fully agree).Social support from adults was measured by the variable Number of trusted adults, that is, how many trusted adults a child had in their life.The variable Number of close friends assessed the respondents' number of close friends by asking: 'Do you currently have close friends, with whom you feel like you can discuss almost everything about your life?'For the sake of interpretation, both social variables were trichotomised ("0 = no close friends/trusted adults, 1 = at least one close friend/trusted adult, 2 = several close friends/trusted adults").
As background variables, we used gender (girl, boy), age and residence (city centre, suburb, rural centre, countryside), family type (one-parent or two-parent family) and parental labour market situation (one working parent vs. two working parent).Initially, we also wanted to include a variable measuring minority affiliation, such as ethnical group affiliation, language group affiliation or disability, since previous research has shown this to be related to stigma (Reutter et al., 2009).However, we had to refrain from doing so, due to the low numbers of respondents belonging to such groups.
Table 1 presents the descriptive statistics for the abovementioned variables in form of number of respondents, means and standard deviations and range as well as confidence intervals for the numerical variables and the percentage distribution for the categorical variables.
Statistically, we initially conducted descriptive statistical analyses (frequency tables and boxplots) to grasp the distribution of variables, and then performed a set of multivariate OLS regression analyses to establish the association between the stigma and poverty variables while simultaneously controlling for other variables.The stigma variables were continuous in these analysis.Stepwise OLS regression was used to see how the variance explained changed by adding new predictors to the model one at a time.The socio-demographic characteristics of the children and their families were first entered into Model 1, including age, gender, family type and parents' labour market situation.Model 2 included well-being and social support variables and final Model 3 included poverty measures. 6The data was analysed using the SPSS 27 statistical software package.

Results
As shown in Table 1, the prevalence of internalized stigma was somewhat higher than experienced stigma.The mean for internalized stigma was 2.17 (standard deviation = 1.00) on a scale from 1 to 5, while the corresponding number for experienced stigma was 1.86 (standard deviation 1.12).This suggests that it is more common to experience feelings that are related to, for example, self-depreciation, shame or guilt than externally perceived feelings of being socially excluded or being slandered by others.
The share of children experiencing subjective poverty and material deprivation largely corresponds to the national child poverty rate in Finland for 2021 (Eurostat, 2023b).While 55.5% of Finnish children reported that their family could make ends meet, about one-third (28.6%) reported some difficulties and 15.9% found it very difficult to make ends meet.As for material deprivation, 18.1% of children were classified as deprived, if one defines deprivation as lacking 2 or more essential items in the deprivation scale (Main, 2013;Main & Bradshaw, 2012).
Next, we turn to our first hypothesis, which postulated that poverty is associated with stigma, that is, that children experiencing subjective poverty and material deprivation also feel more stigmatized.As shown in Fig. 1, which consists of four boxplots graphically displaying how poor children differ from others in terms of experienced and internalized stigma, we see that there is a clear association between experienced and internalized stigma on the one hand and subjective poverty and material deprivation on the other.
The two upper boxplots in Fig. 1 show that children who are subjectively poor are more likely to report both experienced and internalized stigma.In the upper boxplot to the left, we see that the boxes for those children reporting financial difficulties are located higher up in the chart than those reporting no difficulties.The median value for experienced stigma is 1 for those children reporting that it is easy for their families to make ends meet financially, while the median for those reporting little difficulties is 1.75 and those reporting difficulties is 2.5.A test of means (not reported here) shows that the differences are statistically significant (p.< 0.001).We can also see a larger spread of observations for those reporting difficulties than for those saying that it is easy to make ends meet.If we turn to the upper boxplot to the right, we see a similar pattern for internalized stigma, but here the association with subjective poverty is stronger.The median value for internalized stigma is 1.43 for children reporting no difficulties to make ends meet, 2.46 for those reporting some difficulties and 3.12 for those reporting difficulties.We can also note that the spread of observations between value categories are more equal than for experienced stigma.
The two lower boxplots in Fig. 1 show that there is also an association between stigma and material deprivation.If we look at the lower boxplot to the left, we see that the box for materially deprived children are located higher up and is more widely spread that for non-deprived children indicating that deprived children face more experienced stigmatization than non-deprived.The median value for experienced stigma is 1.25 for materially non-deprived children and 2.25 for deprived children.In the lower boxplot to the right, we see a same pattern for internalized stigma, that is, materially non-deprived children report lower internalized stigma than deprived children.The median value for internalized stigma is 1.86 for materially non-deprived children and 2.90 for deprived children.Also here a test of means shows that the reported differences are statistically significant (p.< 0.001).
The results thus far supports H1 by showing that both subjectively poor and materially deprived children report higher experienced and internalized stigma than others, and that this is particularly true for the latter dimension of stigma.The next question is to what extent these bivariate relationships hold if we control for the other relevant variables that have been found to have an impact on stigma, such as social and psychological factors as well as socio-economic factors.Table 2 and 3 report unstandardized Beta coefficients from multivariate OLS regression models for experienced and internalized stigma, respectively.As shown in Table 2, subjective poverty and material deprivation are strong predictors of experienced stigma in Finnish children.Both variables are significantly related to this kind of stigma, but material deprivation seems to play a more important role than subjective poverty.However, we also see that other factors are important for experienced stigma.As postulated in H2, we find that social and psychological factors, such as having friends or having high self-esteem, reduce experienced stigma.Also age and psychological wellbeing matters: older children feel less experienced stigma than younger while low psychological wellbeing increases the risk of it.If we look closer at model 1, we see that age, gender and residence does not predict experienced stigma while family structure (number of adults in the household) and parental labour market situation does, but that these associations become insignificant when controlling for other variables.Model 2 shows that higher self-esteem, good psychological wellbeing, and a higher number of friends reduce the risk of experienced stigma, and that these associations remain significant when controlling for the poverty variables in model 3.
Age also becomes significant in Model 2, suggesting that when children get older, they feel less experienced stigmatization.Finally, Model 3 shows that age, selfesteem and having friends remains significantly associated with experienced stigmatization when all variables are included.We also see that material deprivation and subjective poverty, that is, difficulties to make ends meet, are also significantly associated with experienced stigmatization.Also, there is a significant improvement of the model fit between models 2 and 3, when the variables of poverty were added.
As shown in Table 3, there is a slightly diverging associational pattern for internalized stigmatization and the different independent variables.Also here, the two poverty measures as well as self-esteem and good psychological wellbeing, are significant predictors of stigma, while age and number of friends are not associated with this dimension of stigma.
Subjective poverty is more strongly associated with stigma than material poverty.However, family type and the number of parents in the labour market are significantly associated with internalized stigma, and the latter variable remain significant when controlling for all variables in model 3. Model 2 brings a significant improvement of the model fit suggesting that self-esteem and good psychological wellbeing serve as important barriers to internalized stigma.We can also note that the overall explanative power of these independent variables is higher for internalized stigma (full model adjusted R 2 = 0.348) than for experienced stigma (full model R 2 = 0.225).

Discussion
This article set out to analyse the prevalence of stigma and its relation to poverty among Finnish children with the help of data from the Children's Voice Survey conducted by Save the Children, Finland in 2021.On the basis of our findings, a number of conclusions can be drawn.
Firstly, in line with our first hypothesis, we found a positive correlation between subjective poverty and material deprivation on the one hand, and experienced and internalized stigma on the other, and this association remained significant even when controlling for other relevant variables.This suggests that poor children face a higher risks of feeling stigmatized than other children.However, internalized stigma was somewhat more common among Finnish children than experienced stigma, and also the relationship between the two poverty measures and internalized stigma was stronger than in the case of experienced stigma.The results are similar to those found by Mickelson and Williams' (2008), suggesting that it is more common to experience feelings that are related to, for example, self-depreciation, shame or guilt than externally perceived feelings of being socially excluded or being slandered by others.Our results also support previous research on both the short-term and long-term negative outcomes of poverty (e.g.Danziger & Waldfogel, 2000;Esping-Andersen et al., 2002;Nygård et al., 2019;Schenck-Fontaine & Panico, 2019), suggesting that child poverty should be taken seriously since stigma can lead to serious problems, such as social exclusion, depression or even suicide (Orth & Robins, 2014;Walker, 2014).Subjective poverty had a stronger role for internalized stigma, whereas material deprivation had a stronger connection to experienced stigma.This suggests that material deprivation is something that matters more for outward relations, and thus causes more experienced stigma, whether subjective poverty reflects more squarely inwards.The fact that the amount of friends did not have any significance for internalized stigma can be said to support this interpretation.Then again, feelings and negative self-images may be more sensitive to problems regarding one's family's capacity to make ends meet (cf.Mickelson and Williams, 2008; see also Main & Bradshaw, 2012).
The prevalence of poverty seems to largely correspond to that of the national child poverty rate (e.g.Eurostat, 2023b), both when it comes to subjective poverty and material deprivation.Almost 16% of the children reported that their family experienced difficulties to make ends meet, while 18% reported material deprivation.However, a large share also reported some difficulties, which suggests that financial difficulties among families may be more common, or anchored than this survey suggests (cf.Hakovirta & Kallio, 2015;Hakovirta & Nygård, 2021).It is also likely, that the Covid-19 pandemic may have aggravated the financial situation of families in Finland after 2021, when this survey was conducted (Varanka et al., 2021), which warrants further studies on both poverty and its relation to stigma.
The second conclusion, which is related to our second hypothesis, is that social and psychological factors, and to some extend demographic and socio-economic factors, also play an role for stigma, and can even moderate the relationship between poverty and stigma.Correspondingly, we found that higher self-esteem and good psychological wellbeing reduced both experienced and internalized stigma, while higher age and a greater number of good friends reduced experienced stigma, but not internalized stigma.Self-esteem was more significant for experienced stigma, which is consistent with previous research (cf.Mickelson & Williams, 2008), while the age result is inconsistent and conflicts with the suggestion that older children are more aware of their family's financial situation, which in turn would lead to a higher risk of psychological stress or stigma (Fernqvist, 2013;Hakovirta & Kallio, 2015;Ridge, 2002).We also found that socio-economic factors, such as the number of parents in employment predicted internalized stigma, but not experienced stigma.The 1 3 Poverty as a Driver of Stigma among Finnish Children during… number of good friends was only significant for experienced stigma, which suggest that having such friends serves as a protective factor against stigmatization while at same time it protects from bullying (see Boulton et al., 1999).
Considering that previous studies on poverty and stigma have often been set in more liberal welfare regimes, these results highlight that poverty and material deprivation are also experienced in social-democratic welfare regimes such as Finland, despite generally lower poverty rates and more extensive public service coverage.The factors contributing or protecting against stigmatization among children that are noticeable in this study cannot necessarily be said to be directly affected by the type of welfare regime the children are living in.Further comparison studies would be needed to confidently explore potential differences between liberal and social-democratic welfare regimes.
We can perhaps conclude by saying that experiences of poverty and children's financial situation seem to have a significant connection to feelings of stigma, which in turn has been found to have a connection to social exclusion (e.g.Walker, 2014).It thus seems logical to say that poverty may be an important driver of stigma, but to what extent there is a causal relation between the two is harder to say, since stigma may also lead to poverty (Reeder & Pryor, 2008;Reutter et al., 2009).Nevertheless, this study shows that there is a link between the two, but it also shows that subjective wellbeing, self-esteem and also friends (in the case of experienced stigma), can serve as important protective factors irrespective of the financial situation children live in.
However, there is need to investigate this association since cross-sectional surveys cannot say anything about the causality between poverty and stigma, nor can it say much about how the prevalence of poverty and stigma have changed over time.Therefore longitudinal studies of children's perceptions and experiences are needed, and such studies should also include wider age groups.Furthermore, there are also some other limitations in this study that could be rectified by another sampling method.For instance, due to its character as a convenience sample, the data may be biased, because such samples tend to overly attract mainly individuals with an interest to participate.The data collection method also used voluntary and anonymous participation, which may increase the probability for recruiting individuals who feel strongly about the issue in question and therefore may favour certain outcomes.Furthermore, the distribution of gender was fairly skew as a majority of the respondents were girls, which made it difficult to draw any conclusions based on gender.
Stigma is a social problem that can have severe outcomes on both individuals and society.Previous research shows that children's experiences of stigma may increase during times of stress and hardship such as the Covid-19 pandemic, but also that poverty and a lack of resources can create stigmatization.However, both stigma and poverty are multidimensional phenomena that have been investigated mainly among adults using single indicators.This article contributes to literature by studying children's experiences of stigma and its relation to poverty by using a multidimensional approach.More specifically, it investigates Finnish children's experienced and internalized stigma during the second year of the Covid-19 pandemic, and analyzes how these dimensions are associated with subjective poverty and material deprivation.We use data from the 2021 Children's Voice survey conducted by Save the Children, Finland.The results show that both dimensions of stigma are correlated with subjective poverty as well as material deprivation, even when controlling for socio-economic and other variables.Also low psychological wellbeing is a significant driver of both forms of stigma, while living in a one-parent household was significantly associated with internalized stigma, but not experienced stigma.By contrast, while higher self-esteem seem to reduce both forms of stigma, a higher number of good friends was found to only reduce experienced stigma.As there are both short-and long-term negative outcomes of stigma, for example in terms of mental health problems and social exclusion, child poverty should be taken seriously -especially during times of crises such as pandemics.Funding Open access funding provided by Abo Akademi University (ABO).

Fig. 1 3
Fig. 1 Experienced and internalized stigma by subjective poverty category (upper boxplots) and material deprivation category (lower boxplots) of Finnish children (N=730)

Table 1
Descriptive statistics.Number of respondents (n), means (M), standard deviations (SD), percentages (%), variable range and 95% confidence intervals Poverty as a Driver of Stigma among Finnish Children during…

Table 2
Predictors of children´s experienced stigma (unstandardized OLS regression estimates)Poverty as a Driver of Stigma among Finnish Children during…

Table 3
Predictors of children's internalized stigma (unstandardized regression estimates) Poverty as a Driver of Stigma among Finnish Children during…Author's Contribution The authors confirm contribution to the paper as follows: study conception and design: E. Vuorenlinna, E., Hakovirta, M., Välimäki, S.; data collection: Sarkia, A; analysis and interpretation of results: Vuorenlinna, E., Välimäki, S., Hakovirta, M., Nygård, M., Lindberg, M.; draft manuscript preparation: Vuorenlinna, E., Nygård, M., Lindberg, M. All authors reviewed the results and approved the final version of the manuscript.