Express yourself? Ease to express one’s identity mediates the relationship between national belonging and mental health

ABSTRACT

A number of studies have reported a positive relationship between levels of national identification and well-being. Although this link is clear, the relationship is likely influenced by a number of other variables. In the current study, we examine two such variables: age and the ease with which people feel they can express their identity in the national context. Participants were drawn from three waves (2008–12) of the biannual New Zealand General Social Survey (NZGSS). The NZGSS consists of a number of questions related to well-being. The current study utilised the questions related to one’s sense of belonging to New Zealand, ease to express one’s identity in New Zealand, and mental health. When controlling for physical health, standard of living, and several demographic control variables, there was a clear relationship between one’s sense of belonging to New Zealand and mental health. Further, this relationship was stronger for older than younger participants. Finally, the ease with which participants felt they could express their identity in New Zealand partially mediated the relationship. Future research should elucidate which specific aspects of their identity people feel is constrained in the national context.

KEYWORDS:

• Social identity
• mental health
• identity expression
• identity constraint
• older adults

Introduction

Humans have a propensity to develop deep connections to the places they inhabit (Shumaker and Taylor 1983; Hummon 1992; Hidalgo and Hernandez 2001; Nielsen-Pincus et al. 2010). The strength of these connections is exemplified by the fact that the emotional attachment people form with particular places (i.e. place attachment) often fits the criteria that are used to define the attachment between an infant and their primary caregiver (Bowlby 1969; Ainsworth and Bell 1970). For example, a place can give people an immense feeling of safety and security and, if they were involuntarily removed from that place, the separation may cause them a great deal of emotional distress (Fried 1963). In addition to reflecting the loss of safety and security, peoples emotional response to being removed from a particular place may tap something much deeper, that the place had become part of their identity (Proshansky et al. 1983). Indeed, as Lynd (1958) noted, ‘Some kind of answer to the question Where do I belong? is necessary for an answer to the question Who am I?’ (p. 210).

Many scales of place may come to form part of people’s identity. Although these scales can go well beyond the level of country (e.g. continent), national identity appears to take precedent relative to lower scales such as neighbourhood and town/city (Laczko 2005). Interestingly, this subjective identification with nation confers benefits to well-being (Morrison et al. 2011; Reeskens and Wright 2011; Dimitrova et al. 2013; Zdrenka et al. 2015; Khan et al. 2019). This relationship is consistent with the Social Cure or Social Identity Approach to Health, whereby social identities are hypothesised to play a critical role in the health and well-being of individuals (Haslam et al. 2009; Cruwys et al. 2014; Jetten et al. 2014). Critically, only specific forms of national identification confer benefits to well-being. Reeskens and Wright (2011) draw a distinction between ethnic nationalism and civic nationalism. Ethnic nationalism is authoritarian in nature and tends to focus on enhancing in-group power (DeNeve and Cooper 1998; Sagiv and Schwartz 2000; Schwartz et al. 2000). In contrast, civic nationalism is characterised by openness, trust, and social engagement (Wright 2011; Wright et al. 2012; Reeskens and Wright 2013). Reeskens and Wright (2011) observed that while civic nationalism was positively related to well-being, ethnic nationalism was negatively related.

In the New Zealand context, a number of studies have investigated the contribution of ethnic and civic elements to New Zealand identity or New Zealandness (Thomas and Nikora 1996; Sibley et al. 2011; Humpage and Greaves 2017). For example, Thomas and Nikora (1996) asked secondary school students ‘What things about yourself, and the way you live, do you feel identify you as a New Zealander?’ and ‘When you think of people living in New Zealand, what is it about New Zealanders that makes them different from people in other countries?’. The most common attributes identified were language/accent (46%), Māori culture/being Māori (29%), lifestyle/upbringing (28%) and being born/living in New Zealand (25%). Further, drawing a more clear distinction between ethnic and civic elements, Humpage and Greaves (2017) utilised data from the International Social Survey Programme (ISSP) and demonstrated that civic elements of national identification were the most frequently endorsed items in New Zealand. For example, the most highly endorsed items for ‘truly being a New Zealander’ were feeling like a New Zealander (93%) and respecting New Zealand political institutions and laws (91%). These are construed as civic elements because they indicate that being a New Zealander is not constrained by ascriptive elements (e.g. needing to be born in New Zealand).

Although the endorsement of civic elements is indicative of the type of national identification that confers benefits to well-being, there is little research on the relationship between national identification and well-being in New Zealand (cf. Zdrenka et al. 2015). In the current study, using data from the biannual New Zealand General Social Survey (NZGSS), we investigate the relationship between national identification and mental health. Obviously, the relationship between national identification and mental health is likely influenced by a number of other variables. Taking advantage of some of the broader questions included in the NZGSS, we examine two additional variables, the impact of age and the ease with which people feel they can express their identity in the national context.

Older adults

There is an extensive literature on the increasing importance of place with age (Wiles et al. 2012; Klok et al. 2017; Wiles et al. 2017). Indeed, Rowles (1993) suggested that ‘Our ability to develop and maintain a sense of attachment to place, to sustain a sense of physical, social, and autobiographical insideness … may as we grow older, become increasingly significant in preserving a sense of identity and continuity amidst a changing world’ (p. 66). Although Rowles’ (1993) quote, and the ageing-in-place literature more generally, focus on relatively small scales of place (e.g. home, neighbourhood, and community), many of the arguments hold across scales. Moreover, specific to New Zealand, older participants in the current study grew up at a time when a number of significant national events occurred. For example, unlike younger participants, older participants may have had parents that went to war and the participants themselves grew up in a time when Sir Edmund Hillary conquered Mt Everest (1953), Dame Whina Cooper lead the land march to Parliament (1975), the Waitangi Tribunal was established (1975), there were wide protests against apartheid during the Springbok rugby tour (1981), and New Zealand established itself as a nuclear weapon-free zone (1987). The meaning and connection these events create may influence identity. Thus, in the current study we investigate whether the relationship between national identification and mental health is influenced by age.

Identity expression

Finally, the social psychology literature reveals a number of factors that influence the degree to which people express aspects of their identity (Ellemers 1993). The social identity model of de-individuation effects (SIDE) postulates that identity expression is determined by both a cognitive component, that reflects the salience of an identity in a given social context, and a strategic component, in which people judge whether expressing an identity is appropriate (Spears and Lea 1994; Reicher et al. 1995). For example, when Turkish migrants in the Netherlands were asked to express their identification with both their native and host groups, making both identities cognitively salient, they strategically stressed their dual identity (i.e. being both Turkish and Dutch) more when they thought the audience was Dutch than when they thought it was a Turkish (Barreto et al. 2003). As this example clearly demonstrates, in some contexts people may affirm unique aspects of their identity to a particular audience. In many cases, however, people may also hide an identity that is associated with a stigmatised group or downplay an identity they think the audience will not accept (Goffman 1963; Katz 1981; Tajfel 1981; Jones et al. 1984; Barreto et al. 2006; Morton and Sonnenberg 2011). Strategically, although constraining one’s identity in this way may guard against being a target of prejudice, it may also negatively influence well-being (Barreto et al. 2006; Townsend et al. 2009; Morton and Sonnenberg 2011; Sønderlund et al. 2017). Thus, in the current study we investigate whether identity constraint tempers the relationship between national identity and mental health.

Current study

In the current study, national identity was measured by asking participants the degree to which they feel they belong to New Zealand. The current study had three hypotheses. First, we hypothesised that there would be a positive relationship between national identity and mental health. Second, we hypothesised this relationship would be stronger for older participants. Finally, we hypothesised that identity constraint would temper the relationship between national identity and mental health. To address these questions, we draw data from three waves (2008–12) of the NZGSS. Broadly, the NZGSS provides information on the well-being of New Zealanders aged 15 years and over and includes questions on a range of social and economic outcomes.

Method

Participants

Participants were drawn from three waves (2008–12) of the biannual NZGSS. The NZGSS consists of a number of questions related to well-being. The sampling strategy is designed to assure a fair chance for every household in the country to be selected and the weighted NZGSS offers a representative sample of New Zealand’s population at each wave (StatsNZ 2019). More specifically, the NZGSS uses a three-stage sample selection method. In Stage 1, primary sampling units (PSUs) are selected from the Household Survey Frame (HSF). The HSF lists PSUs with attributes determined by data from the national census. In Stage 2, eligible dwellings are selected. Finally, in Stage 3 one eligible individual within each dwelling is chosen at random from all eligible individuals in that dwelling. The NZGSS is designed to provide estimates at a national level. The response rates for the 2008, 2010, and 2012 waves were 83%, 81%, and 78%, respectively. Excluding the responses of individuals who did not complete all of the questions of interest in the present study, our total sample was 24,328. A more complete description of the sample’s characteristics is presented in Table 1.

Table 1. Descriptive statistics^a.

NZGSS Wave (Year)	N	Missing values %	Gender (female %)	Ethnic group %	Age group	Education	Annual personal income	Social marital status (partnered %)	Self-rated standard of living	Overall physical health status	Overall mental health status	Sense of belonging to NZ	Ease to express one’s identity in NZ
1st Wave (2008)	8158	6.5	55.2	European: 77.0 Māori: 6.1 Pacific: 3.1 Asian: 5.6 European-Māori: 4.2 Others: 3.9	7.51 (3.58)	3.62 (3.00)	18.81 (3.65)	57.8	12.06 (0.86)	49.74 (10.67)	52.24 (9.19)	2.39 (0.79)	4.18 (0.80)
2nd Wave (2010)	8118	5.1	56.2	European: 76.7 Māori: 6.4 Pacific: 2.4 Asian: 6.1 European-Māori: 4.1 Others: 4.3	7.65 (3.59)	3.74 (2.99)	18.65 (3.62)	57.5	12.50 (0.82)	48.64 (10.73)	49.46 (10.82)	2.39 (0.79)	4.20 (0.80)
3rd Wave (2012)	8052	4.8	56.5	European: 72.4 Māori: 7.7 Pacific: 3.6 Asian: 7.0 European-Māori: 4.6 Others: 4.5	7.59 (3.60)	3.69 (2.96)	18.77 (3.61)	56.0	12.52 (0.84)	48.41 (10.87)	49.69 (10.71)	2.37 (0.79)	4.20 (0.79)
Total Sample	24328	5.5	56.0	European: 75.4 Māori: 6.7 Pacific: 3.1 Asian: 6.2 European-Māori: 4.3 Others: 4.2	7.58 (3.59)	3.69 (2.99)	18.74 (3.63)	57.1	12.36 (0.86)	48.93 (10.77)	50.47 (10.34)	2.38 (0.79)	4.19 (0.80)

^aStandard deviations have been shown between the brackets.

Measures

Sense of belonging to New Zealand

Two consecutive questions in the NZGSS concerned sense of belonging to New Zealand: (1) ‘do you feel that you belong to New Zealand?’ (1 = yes, 95.1% in 2008, 94.7% in 2010, and 94.4% in 2012, and 2 = no), and (2) ‘would you say you feel that very strongly, strongly or not very strongly?’ (5.7% in 2008, 5.8% in 2010, and 5.3% in 2012 responded ‘not very strongly’, 39.5% in 2008, 39.5% in 2010, and 41.8% in 2012 responded ‘strongly’, and 54.8% in 2008, 54.7% in 2010, and 52.9% in 2012 responded ‘very strongly’). Only participants that answered ‘yes’ to the first question were asked the second question. Given this, we collapsed the questions such that participants that answered no to the first question were recoded as 0 and, for those that answered yes to the first question, their response to the second question was recoded as 1 (‘not very strongly’), 2 (‘strongly’), or 3 (‘very strongly’). Recoding responses in this manner provided us with a 4-point likert scale. Missing values in each wave were 1.28% (2008), 0.77% (2010), and 0.65% (2012).

Ease to express one’s identity in New Zealand

A single item assessed the ease with which individuals felt they could express their identity in New Zealand (i.e. ‘here in New Zealand, how easy or difficult is it for you to express your own identity?’). Participants responded using a scale from 1 (‘very easy’) to 5 (‘very difficult’). We recoded the original responses so that high scores represented easier expression of one’s identity in New Zealand. Missing values in each wave were 0.15% (2008), 0.18% (2010), and 0.18% (2012).

Overall mental health status

Mental health status was derived from the Mental Component Summary (MCS) of the 12-item Short-Form (SF-12) Health Survey (Ware et al. 1996). The MCS includes items assessing feeling/mood (e.g. peaceful, depressed, downhearted, etc.). Items included ‘how much of the time during the past four weeks have you felt downhearted and depressed?’ and ‘accomplished less than you would like as a result of any emotional problems, such as feeling depressed or anxious?’ Scores were rescaled to a score between 0–100 with high scores indicating enhanced overall mental health status. Critically, previous research has demonstrated the reliability and validity of the SF-12 across a wide range of groups and nationalities (Gandek et al. 1998; Jenkinson et al. 2001; Andrews 2002).

Controls

Participant’s physical health status was assessed using the Physical Component Summary (PCS) of the SF-12 (Ware et al. 1996) and rescaled to a score between 0–100 with higher scores indicating better overall physical health status. The PCS includes items assessing general health and activity level. Self-rated standard of living was assessed using a single item (‘generally, how would you rate your standard of living?’), with participants responding on a 5-point scale (1 = high; 5 = low). Finally, we included several demographic control variables. Specifically, age group (fourteen 5-year intervals with a final interval of 85 and above, e.g. 1st interval: 15–19, 2nd interval: 20–24, etc.), gender (male/female), ethnicity (recoded into six major ethnic groups, i.e. New Zealand European, Māori, Pacific, Asian, European-Māori, and Others), education (coded between 0–10 representing a range of ‘no qualification’ to ‘PhD’), marital status (partnered/non-partnered), and annual personal income (sixteen intervals from loss to 150 K and above).

Analytic strategy

The NZGSS data are time series cross-sectional data, with individual participant’s data nested within waves. Although our hypotheses only concern the relationship between individual-level variables, our statistical analyses take the multilevel structure of the data into account. To this end, we employed both fixed effects (FEs) and random effects (REs) models. In the FE models, unit dummies (i.e. one dummy for each wave except for a reference wave) were used to account for by-class (i.e. NZGSS wave) heterogeneity in the outcome variable (i.e. overall mental health status). In the RE models, a random intercept was used for this purpose. Whether FE or RE models are more appropriate in our case is subject to an ongoing debate in the literature (Bryan and Jenkins 2015; Sortheix et al. 2017). Nevertheless, there is a general agreement that when there are a very few classes (e.g. three NZGSS waves), class-specific estimates in RE models can be biased (Bryan and Jenkins 2015; Hox et al. 2017). Taking a conservative approach, we perform the analyses using both FE and RE models and only interpret results that are significant in both models. The results of the FE models are presented here (Table 2) and the results from the RE models are presented in the Online Supplementary Material (Tables S1 and S2). FE model analyses were carried out using SPSS and the PROCESS macro (Hayes 2013) and the RE model analyses were conducted using R and the R-packages lme4 (Bates et al. 2015) and mediation (Tingley et al. 2014).

Table 2. Results for analyses regressing sense of belonging to New Zealand and control variables on ease to express one’s identity in New Zealand and overall mental health status with NZGSS wave FEs.

		Model 1 (DV: Overall mental health status)		Model 2 (DV: Overall mental health status)		Model 3 (DV: Overall mental health status)		Model 4 (DV: Ease to express one’s identity in NZ^a)		Model 5 (DV: Overall mental health status)
		B	SE	B	SE	B	SE	B	SE	B	SE
Sense of belonging to NZ^a		.82***	.07	.55***	.07	.57***	.07	.20***	.01	.17*	.07
NZGSS wave	2010 dummy	−1.32***	.08	−.66***	.07	−.66***	.07	.04***	.01	−.73***	.07
	2012 dummy	−1.20***	.08	−.52***	.07	−.52***	.07	.04***	.01	−.60***	.07
Age group				.75***	.07	.73***	.07	.05***	.01	.66***	.07
Gender				−.55***	.07	−.54***	.07	.00	.01	−.55***	.06
Ethnicity	Māori dummy			−.13*	.07	−.13^†	.07	.01**	.01	−.16*	.07
	Pacific dummy			.25***	.07	.25***	.07	.01	.01	.24**	.07
	Asian dummy			.59***	.08	.57***	.08	−.06***	.01	.71***	.08
	European−Māori dummy			−.16*	.07	−.16*	.07	−.01	.01	−.15*	.07
	Others dummy			−.14*	.07	−.14*	.07	−.01**	.01	−.11^†	.07
Education				−.13^†	.07	−.13^†	.07	.04***	.01	−.20**	.07
Social marital status				−.76***	.06	−.76***	.06	.00	.00	−.76***	.06
Annual personal income				.59***	.08	.60***	.08	.02**	.01	.56***	.07
Overall physical health status				−.10	.06	−.10	.06	.02***	.00	−.13*	.06
Self−rated standard of living				−2.72***	.07	−2.71***	.07	−.13***	.01	−2.48***	.07
Sense of belonging to NZ^a × Age group						.21**	.07
Ease to express one’s identity in NZ^a										1.51***	.06
Overall R^2		.02		.12		.12		.11		.25
Overall F		173.2***		214.8***		202.0***		206.5***		239.4***
Df		24,324		24,312		24,311		24,312		24311

Notes. N = 24,328. The table presents the unstandardised b-coefficients and standard errors for centred variables.

^aNew Zealand. *p < .05. **p < .01. ***p < .001. ^†p < .10.

Results

Table 2 presents the results of the FE models regressing sense of belonging to New Zealand on overall mental health status. Model 1 only included the FEs for sense of belonging to New Zealand and NZGSS wave. Sense of belonging to New Zealand had a statistically significant effect in the FE (see Table 2, Model 1) and RE models (Online Supplementary Table S1, Model 1) on overall mental health status. Critically, the relationship between sense of belonging to New Zealand and overall mental health status remained significant when controlling for age group, gender, ethnicity, education, annual personal income, social/marital status, overall physical health status, and self-rated standard of living (see Table 2, Model 2 and Online Supplementary Table S1, Model 2).

To investigate the effect of age, we extended Model 2 of Table 2 (and Model 2 of Online Supplementary Table S1) by adding an interaction term for sense of belonging to New Zealand and age group. The results showed that the interaction effect of sense of belonging to New Zealand and age group was significantly associated with overall mental health status (see Table 2, Model 3 and Online Supplementary Table S1, Model 3). As shown in Figure 1, the relationship between sense of belonging to New Zealand and overall mental health status was stronger for older than younger participants. Although stronger for older participants, a simple slope analysis confirmed that the positive association between sense of belonging to New Zealand and overall mental health status held for both younger (simple slope B = .37, t (24,311) = 4.26, p < .0001, 95% CI [0.20, 0.55], for one SD below the mean age group) and older participants (simple slope B = .78, t (24,311) = 7.70, p < .0001, 95% CI [0.58, 0.98], for one SD above the mean age group).

Figure 1. Interaction between sense of belonging to New Zealand and age group on overall mental health status. The scale ranges from 0 to 100 with high scores indicating enhanced overall mental health status.

The mediating role of identity expression

We examined the role of ease to express one’s identity in New Zealand as a mediator in the relationship between sense of belonging to New Zealand and overall mental health status. Using the PROCESS plug-in for SPSS with 1,000 resamples for the FE model (and the mediation package in R with 1,000 Monte Carlo draws for nonparametric bootstrap for the RE model), a mediator model that controlled for the covariates was utilised to test the indirect path between sense of belonging to New Zealand and overall mental health status. Fittingly, the model supports that both the direct effect (estimate = .17, p = .012, 95% CI [0.04, 0.30]) and the indirect effect (estimate = .38, 95% CI [0.34, 0.42]) of sense of belonging to New Zealand on overall mental health status were statistically significant (ratio of the indirect effect to the total effect estimate = .69, 95% CI [0.53, 0.94]). Similarly, the RE models mediation analysis yielded that the average direct effect (estimate = .17, p < .0001, 95% CI [0.04, 0.30]) and the average indirect effect (estimate = .30, p < .0001, 95% CI [0.27, 0.33]) of sense of belonging to New Zealand on overall mental health status was statistically significant (proportion mediated estimate = .64, p < .0001, 95% CI [0.49, 0.89]). That is, ease to express one’s identity in New Zealand partially mediated the effect of sense of belonging to New Zealand on overall mental health status (also see Table 2, Models 4 and 5 and Online Supplementary Table S2, Models 1 and 2).

Although the surveys did not include a question asking about the aspects of their identity (e.g. ethnicity, sexuality, etc.) participants were thinking about when answering the question on identity expression, there was a follow-up question regarding the reasons people find it difficult to express their identity. Specifically, participants that responded ‘sometimes easy/sometimes difficult’ (3), ‘difficult’ (4), or ‘very difficult’ (5) to the identity expression question were presented with a single item (‘what things make it difficult for you?’) and the following response options: ‘it’s illegal to do it’, ‘there’s no place to do it’, ‘some people won’t accept it’, ‘I worry about what other people would think’, ‘I worry about what other people might do’, ‘other’, ‘don’t know’, and ‘refused’. Participants were allowed to select multiple response options. As shown in Figure 2, across all three survey waves the options ‘some people won’t accept it’ and ‘I worry about what other people would think’ were the most commonly cited reasons for participants not feeling they could express their identity in New Zealand.

Figure 2. Participant’s responses to the question (‘what things make it difficult for you?’) regarding what makes it difficult for them to express themselves in New Zealand.

Discussion

In the current study, we investigated the relationship between national identity (i.e. sense of belonging to New Zealand) and mental health, utilising data from three waves (2008–12) of the biannual NZGSS. Consistent with our first hypothesis, and earlier work in this area, there was a positive relationship between national identity and mental health (Morrison et al. 2011; Reeskens and Wright 2011; Dimitrova et al. 2013; Zdrenka et al. 2015). Consistent with our second hypotheses, the relationship between national identity and mental health was stronger for older than younger participants. Finally, consistent with our third hypothesis, the ease to express one’s identity in New Zealand partially mediated the relationship between national identity and mental health.

As noted above, the stronger relationship between national identity and mental health for older participants may be due to the increasing importance of place with age (Wiles et al. 2012; Klok et al. 2017; Wiles et al. 2017) and/or cohort effects related to the fact older participants grew up in New Zealand at a time when a number of significant national events occurred. Moreover, many of these events (e.g. the establishment of the Waitangi Tribunal, establishing New Zealand as a nuclear weapon-free zone) likely shaped the New Zealand identity (Liu et al. 2005).

With respect to identity expression, consistent with the strategic component of the SIDE model, the most common reasons for not expressing identity were due to concerns about acceptance and worrying about what other people would think (Spears and Lea 1994; Reicher et al. 1995). A number of theories may explain the identity expression finding. First, the inability to express an important identity may compromise an individual’s sense of self (Morton and Sonnenberg 2011). Indeed, the self-concept is not only informed by how people see themselves but also how others see them, and not expressing an important identity may create a conflict between these two perspectives (Barreto and Ellemers 2003; Morton and Sonnenberg 2011). Second, while not expressing an important identity may protect one from prejudice, it comes at the cost of not having that identity recognised, respected, and accepted by others (Swann et al. 1989; Barreto and Ellemers 2002; Ryan et al. 2010). While each of these acts by others (i.e. recognition, respect, and acceptance) is important, as participants reasons for not expressing their identity suggest, acceptance is at the forefront of many people’s minds. Indeed, acceptance helps to satisfy core social motives like belonging, control, self-esteem, and meaning (Baumeister and Leary 1995; Fiske 2009; Williams 2009) which, in turn, have been linked to multiple aspects of mental health (Eisenberg and Resnick 2006; DiFulvio 2011; Greenaway et al. 2016; Scarf et al. 2016; Scarf et al. 2017; Arahanga-Doyle et al. 2018; Scarf et al. 2018; Koni et al. 2019). Thus, while not expressing their identity in New Zealand may protect people from being the target of prejudice, it also robs them of the many benefits acceptance brings (Baumeister and Leary 1995).

The identity expression finding may also have implications for multiculturalism in New Zealand. We did find reliable differences in mental health between ethnic groups with, for example, Māori participants reporting somewhat worse mental health than New Zealand European participants and, conversely, Asian participants reporting better mental health (see Table 2, Model 2). However, ease of identity expression does not appear to account for these differences because it is actually slightly higher for Māori and markedly lower for Asian respondents. Levels of national belonging followed a similar but less pronounced pattern with Māori participants reporting slightly more belonging, and Asian (and to a lesser extent Pacific) respondents reporting slightly less. The relationship between these variables needs further investigation given that the current dataset does not allow us to confirm which identities participants had in mind when responding to this question. One might predict that the accommodation of difference entailed my multiculturalism has important health implications because of the way it enables both national belonging and identity expression among minority groups.

Limitations

The current study is not without limitations. First, the findings are based on cross-sectional data so causation cannot be established. Although national identity may contribute to mental health, it is also possible that better mental health may lead one to feel a greater sense of belonging to New Zealand. Most likely, however, is that the relationship is bidirectional (Saeri et al. 2018; Khan et al. 2019). Second, place identity was measured using a single item that asked about participant’s sense of belonging to New Zealand. Belonging is a central component in definitions of place identity (Hernández et al. 2007) and social identity (Scarf et al. 2016). Further, Korpela (1989) described belonging as ‘ … not only one aspect of place-identity, but a necessary basis for it’ (p. 246). However, place identity is likely multidimensional and a comprehensive measure would allow one to determine whether it is belonging specifically, or place identity more generally, that relates to mental health (Raymond et al. 2010). Finally, as noted above, the identity constraint measure was limited to a single item that assessed the ease with which individuals felt they could express their identity in New Zealand (i.e. ‘here in New Zealand, how easy or difficult is it for you to express your own identity?’). Future studies should utilise identity constraint measures that (1) give participants the option to complete identity constraint measures for each identity they believe is constrained and, (2) distinguish between a person’s perceived freedom to express an identity and the degree to which they believe others accurately see them. For example, Sønderlund et al. (2017) perceived identity expression scale includes an item assessing freedom to express an identity (i.e. ‘In general, I feel free to fully express myself and my identity to the people around me’) and two reverse scored items assessing identity recognition (e.g. ‘Other people don’t see me the way I want to be seen’).

Conclusion

On a positive note, the current study shows a clear relationship between place identity at the national level and mental health. However, our results also demonstrate that this relationship is partially mediated by the ease with which one feels they can express their identity in New Zealand. Although we can only speculate about the specific aspects of their identity that people feel will not be accepted or might be judged, it appears there is still work to be done in challenging stigma and discrimination in New Zealand.

Acknowledgements

Access to the data used in this study was provided by Statistics New Zealand under conditions designed to keep individual information secure in accordance with requirements of the Statistics Act 1975. The opinions presented are those of the author(s) and do not necessarily represent an official view of Statistics New Zealand.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Saleh Moradi http://orcid.org/0000-0003-0668-5891

Hitaua Arahanga-Doyle http://orcid.org/0000-0002-5416-1008

Damian Scarf http://orcid.org/0000-0002-9486-8059