The Australian Women's Loneliness Report

What the data tells us

A curated synthesis of existing Australian research on women's loneliness and social isolation, prepared by Safe Refuge Counselling, a private online and in-person counselling practice for women based in Mount Barker, South Australia. It is not a domestic violence service, despite what the name can suggest at first reading.

Content note: this report discusses loneliness, mental health, and psychological distress. Help-seeking resources are listed at the end.

---

1. Executive summary

Loneliness in Australia is not evenly distributed. The existing public evidence base, drawn from the Australian Institute of Health and Welfare (AIHW), the Household, Income and Labour Dynamics in Australia (HILDA) survey, the Australian Bureau of Statistics (ABS), and Ending Loneliness Together, shows that women carry a specific and sustained loneliness burden across the life course. The drivers shift with life stage and context, but the pattern is consistent: women report more loneliness than men in most age bands, and women in particular circumstances (caring, migrating, ageing, parenting young children, living rurally, living with disability) sit above the already elevated national baseline.

This report is a synthesis, not new research. Every number below is cited to a specific public source so a reader can verify it.

Headline statistics

1. In 2023, an estimated 15 per cent of Australians experienced loneliness on the UCLA loneliness scale in HILDA, with women slightly more affected than men (16 per cent of females, 15 per cent of males). Source: AIHW, Australia's Welfare 2023, Chapter 2.
2. Approximately one in three Australian adults (32 per cent) reported being lonely in the 2023 State of the Nation Report on Social Connection, with one in six (17 per cent) reporting severe loneliness. Women's rate (32 per cent) was marginally higher than men's (31 per cent). Source: Ending Loneliness Together, State of the Nation 2023.
3. Using HILDA data from 2014 to 2018, 13 per cent of Australians met criteria for chronic loneliness and a further 21 per cent for episodic loneliness, meaning about one in three adults experienced meaningful loneliness across that five-year window. Source: Lim et al., Scientific Reports (2023).
4. Three in ten (29 per cent) people with disability aged 15 to 64 say they often feel lonely, compared with 17 per cent of those without disability. Source: AIHW, People with Disability in Australia.
5. Among women aged 75 and over, 40 per cent live alone, compared with 22 per cent of men the same age, and the prevalence of loneliness peaks at 19 per cent for people aged 75 and over. Source: ABS 2021 Census; AIHW analysis.

These numbers are not a verdict on individual women. Loneliness is a signal, not a diagnosis. Most women who feel lonely are responding, understandably, to real circumstances (a move, a birth, a loss, a caring role, a health event) rather than to any personal deficit. The evidence base supports that framing.

---

2. Why loneliness is a women's issue

Loneliness is often discussed as if it arrives out of nowhere. The Australian data suggest otherwise. Three structural patterns explain why women, in aggregate, report more loneliness than men.

Women live longer and more often alone in later life. ABS 2021 Census data show 40 per cent of women aged 75 and over live in lone-person households, almost double the proportion of men at the same age (22 per cent). Living alone is one of the strongest demographic correlates of loneliness in Australian research. Source: ABS 2021 Census; AIHW Older Australians report.

Women carry more unpaid care. The ABS General Social Survey and time-use data consistently show women doing more of the unpaid caring (children, ageing parents, partners with illness). Intensive caring roles reduce discretionary time for social contact. Source: ABS, General Social Survey: Summary Results 2020.

Women experience specific life-course shocks to social networks. Matrescence, perinatal mental health difficulties, relationship breakdown, widowhood, menopause, and caring responsibilities all cluster disproportionately on women. Each can thin a social network at the exact moment more support is needed. Source: Ending Loneliness Together, Why We Feel Lonely (2024).

[CHART: Horizontal bar chart. Title: "Self-reported loneliness prevalence by gender, Australia 2023." Series: Female 16%, Male 15%. X-axis: percentage of respondents. Source: AIHW, Australia's Welfare 2023, drawing on HILDA 2023.]

None of this is about women being more fragile. It is about where the loneliness load sits in a country that still defaults to women for care and connection work. Naming that pattern is the first step to responding to it without shame.

---

3. Loneliness by life stage

3.1 Adolescent and young adult women (15 to 24)

Young women are the life-stage group showing the steepest loneliness trajectory in Australian data. AIHW analysis of HILDA shows an increasing trend in loneliness among 15 to 24 year olds, especially females, since 2012. Mean loneliness scores for women aged 15 to 24 (3.3) have exceeded those for men the same age (3.0) in recent waves. Source: AIHW, Australia's Welfare 2023, Chapter 2, citing HILDA.

Ending Loneliness Together's State of the Nation data indicate that 18 to 24 year olds report loneliness at roughly four times the rate of Australians aged 75 and over. Source: Ending Loneliness Together, State of the Nation 2023.

[CHART: Line chart. Title: "Mean loneliness score, Australians aged 15 to 24, by gender, 2001 to 2023." Two lines: females and males. Y-axis: mean UCLA loneliness score. X-axis: survey wave. Source: HILDA, as reported in AIHW Australia's Welfare 2023.]

What the data do not tell us is why. Plausible contributors named in the NSW Loneliness Inquiry submissions include housing insecurity, study-related pressure, social media use, and the delayed partnering and parenting patterns of this generation. The evidence does not support a single causal claim.

3.2 Early career and partnering years (25 to 34)

HILDA data show loneliness for women aged 25 to 54 consistently exceeding men's in the same band, a pattern observed in every wave between 2001 and 2023. Source: AIHW, Australia's Welfare 2023.

This is the life stage of first-home-buying, relationship formation, relocation for work, and often the first significant drop-off from established friendship groups. The data show the loneliness, but cannot isolate which of these pressures drives it.

3.3 Matrescence and early motherhood

Matrescence, the developmental transition into motherhood, is a predictable period of social reorganisation. International research cited by Australian perinatal services suggests that a majority of new parents experience meaningful loneliness. A scoping review published in 2021 summarised multiple studies showing high rates of loneliness in pregnant women and mothers of young children. Source: Nowland et al., Perspectives in Public Health (2021), and BMC Systematic Reviews scoping review (2022).

Perinatal Anxiety and Depression Australia (PANDA) estimates that up to one in five women will experience anxiety or depression during the perinatal period, affecting approximately 100,000 Australian families each year. Loneliness is a well-documented correlate of perinatal mental health difficulty, though not a cause in itself. Source: PANDA, How common are mental health issues in the perinatal period?

[CHART: Stacked bar. Title: "Perinatal mental health and social isolation co-occurrence, Australia." Series: estimated 1 in 5 women perinatal anxiety/depression; high overlap with reported loneliness. Note: data are estimates from PANDA and peer-reviewed scoping reviews, not a single unified dataset. Source: PANDA; Nowland et al. 2021.]

3.4 Mid-life women (45 to 64)

The Australian Longitudinal Study on Women's Health has followed Australian women for more than 25 years. A 2025 analysis identified longitudinal patterns of loneliness over 18 years in middle-aged women, finding that women with 'stable-high' or 'increasing' loneliness trajectories had substantially lower health-related quality of life than women with stable-low loneliness. Baseline predictors of sustained loneliness included depression, anxiety, stress, smoking, and low social support. Source: Communications Psychology (2025), analysis of ALSWH data.

Mid-life is also when many women manage 'sandwich' caring (children still at home, parents declining), divorce or separation, menopause, and career transition. These are load factors the evidence base treats as correlates, not causes.

3.5 Older women (65 and over)

The AIHW reports national prevalence of loneliness of approximately 17 per cent, rising to around 19 per cent for Australians aged 75 and over. Women in this age band are disproportionately likely to live alone: 40 per cent of women aged 75+ live alone, compared with 22 per cent of men. Source: AIHW; ABS 2021 Census.

Living alone does not equal being lonely. The Australian evidence consistently finds that quality of contact, not household composition, predicts loneliness most strongly. However, living alone is associated with higher odds of loneliness (odds ratio approximately 2.86 in retirement-living samples). Source: BMC Geriatrics cross-sectional study, 2025.

[CHART: Grouped bar chart. Title: "Women aged 75 and over: living arrangement and loneliness, Australia." Bars: % living alone (women 40%, men 22%); % lonely by age 75+ (approx. 19%). Source: ABS 2021 Census; AIHW.]

---

4. Loneliness by context

4.1 Rural and remote women

A 2024 web-based cross-sectional survey reported that loneliness affects approximately 35 per cent of rural Australians. Source: SSM - Population Health (2024), 'Factors associated with loneliness in rural Australia.'

The National Rural Health Alliance notes that Australians living rurally face a combination of isolation, financial pressure, and natural-disaster exposure that compounds mental health risk. Evidence specifically on farming women's loneliness is thinner than the equivalent body of research on farming men, a gap recognised in recent peer-reviewed literature. Source: National Rural Health Alliance, Mental Health in Rural and Remote Australia fact sheet (2021); Bryant, Sociologia Ruralis (2026, early online).

A 2024 study of perinatal mental health and rurality in an Australian cohort found an association between rurality and perinatal distress, consistent with the broader pattern that geographic isolation amplifies already-elevated perinatal risk. Source: Rural and Remote Health, 2024.

4.2 Culturally and linguistically diverse (CALD) women

Australian research on loneliness specifically among CALD women is limited but consistent in direction. Submissions to the NSW Loneliness Inquiry from organisations including the Multicultural Communities Council of NSW identified language, digital literacy, small social networks, and limited prior exposure to Australian civic life as barriers to connection. Source: NSW Parliament, Standing Committee on Social Issues, Inquiry into the prevalence, causes and impacts of loneliness in NSW, Submission 78.

A 2023 scoping review on CALD ageing experiences found that migration history, language, and cultural identity interact with ageing-related loneliness in ways the general population data do not capture. Source: Radermacher et al., The Gerontologist (2023).

4.3 Aboriginal and Torres Strait Islander women

Loneliness data specific to Aboriginal and Torres Strait Islander women must be read alongside the broader framework of social and emotional wellbeing, which treats connection to family, community, Country, and culture as foundational. Source: AIHW, Indigenous Health Performance Framework, Measure 1.18.

The National Aboriginal and Torres Strait Islander Health Survey (2018 to 2019) found that participants who were removed, or whose relatives had been removed, from their family were more likely to have high or very high levels of psychological distress (38 per cent) than those who had not experienced removal (26 per cent). Source: ABS, National Aboriginal and Torres Strait Islander Health Survey, 2018 to 2019.

The National Mental Health Commission's 2024 report on the mental health and social and emotional wellbeing of Aboriginal and Torres Strait Islander peoples emphasises that loneliness interventions not grounded in community, kinship, and cultural continuity are unlikely to work. Source: National Mental Health Commission (2024).

4.4 Women with disability

The AIHW reports that 3 in 10 (29 per cent) of people with disability aged 15 to 64 often feel lonely, compared with 17 per cent of those without disability. One in five (19 per cent) also experience social isolation, compared with 9.5 per cent of those without disability. Source: AIHW, People with Disability in Australia, 2021 data.

A peer-reviewed analysis of Australian trends from 2003 to 2020 found that loneliness prevalence for people without disability declined over the period, but loneliness prevalence for people with disability did not. Source: Aitken et al., BMC Public Health (2024).

[CHART: Grouped bar chart. Title: "Loneliness and social isolation: people aged 15 to 64 with and without disability, Australia 2021." Bars: often lonely (29% with disability vs 17% without); socially isolated (19% vs 9.5%). Source: AIHW.]

4.5 LGBTQIA+ women

LGBTQIA+ Australians report lower perceived social support and higher risk of social isolation than heterosexual and cisgender peers. Source: Lim et al., peer-reviewed analysis in Psychology & Sexuality (2020), cited via PubMed ID 33315419.

Research specifically with older lesbian women aged 60 and over found that being single and having less connection to lesbian and gay communities predicted higher loneliness, and that more frequent recent experiences of sexual orientation discrimination predicted loneliness for women. Source: Lyons et al., Journal of Homosexuality / Aging & Mental Health (2021).

Intersectional caveat

Very little Australian data disaggregates to intersectional subgroups (for example, Aboriginal LGBTQIA+ women, CALD women with disability). Most headline statistics quoted above are single-axis. Reports that claim intersectional numbers without a cited source should be read cautiously.

---

5. Loneliness and mental health outcomes

The relationship between loneliness and mental health is well established in Australian data, but the direction of causation is not. The honest summary is: loneliness and poor mental health travel together, each can precede and intensify the other, and reducing loneliness is a legitimate mental health strategy even when the causal pathway cannot be proven for an individual.

Psychological distress. HILDA analysis shows females categorised as lonely report significantly higher psychological distress. Mean loneliness scores for women with poor mental health (4.2) are substantially higher than for women without poor mental health (2.5). Source: AIHW, Australia's Welfare 2023, citing HILDA.

Depression risk. A widely cited meta-analysis reports that adults who are often lonely have a pooled adjusted odds ratio of approximately 2.33 for new-onset depression, compared with those who are not often lonely. Source: Mann et al., Social Psychiatry and Psychiatric Epidemiology (2022), as cited in Australian policy documents.

Anxiety. Australia's 2020 to 2022 National Study of Mental Health and Wellbeing found 21.1 per cent of women had a 12-month anxiety disorder, compared with 13.3 per cent of men. Source: ABS, National Study of Mental Health and Wellbeing 2020 to 2022.

Chronic disease comorbidity. The Ending Loneliness Together State of the Nation 2023 report estimates lonely Australians are approximately 4.6 times more likely to report depression and about twice as likely to report chronic disease than non-lonely Australians. These are associations drawn from cross-sectional survey data and should not be read as causal. Source: Ending Loneliness Together, State of the Nation 2023.

Persistence effects. People with mental health conditions such as social anxiety and depression are 2.9 times more likely to experience persistent loneliness than those without mental health conditions. Source: Ending Loneliness Together, Why We Feel Lonely (2024).

[CHART: Horizontal bar chart. Title: "Odds ratios and relative risks for mental health conditions in lonely vs non-lonely Australians." Bars: 4.6x depression (State of the Nation 2023); 2.9x persistent loneliness in people with mental health conditions (Why We Feel Lonely 2024); 2.33 meta-analytic OR new-onset depression (Mann et al. 2022). Source: as labelled. Note: figures from different study designs and not directly comparable.]

A note on framing, consistent with Mindframe guidelines. Loneliness is a human response to a gap between the connection a person has and the connection they want. It is not a character flaw. It is not a life sentence. It does not make a woman dangerous to herself or others. It is, for most women, modifiable.

---

6. What the evidence says about interventions that work

The intervention evidence base is still uneven. The 2024 Campbell systematic evidence-and-gap map of in-person loneliness interventions found many published trials but only a small proportion at high or moderate methodological quality. Source: Welch et al., Campbell Systematic Reviews (2024).

Within that caveat, four categories have emerging Australian evidence.

1. Group-based interventions grounded in social identity. The Groups for Belonging intervention, trialled in Australia, uses a social-identity approach with psychoeducation to reduce loneliness in people attending alcohol and other drug treatment. Source: BMC Public Health (2025), cluster randomised controlled trial.

2. Social prescribing. An 8-week controlled evaluation in Queensland found social prescribing participants reported improvements in loneliness and social trust, while usual-care controls did not. Source: Frontiers in Psychology (2024).

3. Community-connection programs. A South Australian community-connections pilot found significant decreases in loneliness and improvements in quality of life at program completion. Source: PLOS One / BMC Public Health (2024), South Australian community intervention pilot.

4. Befriending. Peer and volunteer befriending shows promising but methodologically variable effects. Source: pilot studies cited in Clinical Gerontologist (2025).

What is not yet proven. Digital-only interventions, single-session workshops, and campaign-only responses have not demonstrated sustained loneliness reduction in peer-reviewed Australian evaluations to date.

What the NSW Loneliness Inquiry recommended. The Standing Committee on Social Issues tabled its final report in 2025, recommending a coordinated NSW response to loneliness, with specific attention to people with disability, single parents, people with financial insecurity, and people with mental health conditions. Source: NSW Parliament, Prevalence, causes and impacts of loneliness in New South Wales, Report No. 65.

[CHART: Table. Title: "Evidence base for loneliness interventions in Australia." Columns: intervention type, evidence level, representative study. Rows: group-based social-identity (RCT, Groups for Belonging); social prescribing (controlled evaluation, Qld); community connections (pilot, SA); befriending (pilot studies, residential aged care). Source: as labelled.]

---

7. If you are a woman who feels lonely. A short guide, in Aana's voice

This section is written by Aana Carpenter, registered counsellor (ACA Level 1), founder of Safe Refuge Counselling, a private online and in-person counselling practice for women.

If you have read this far, you may be recognising something of your own experience in the data. That recognition itself is useful information. Loneliness thrives on the belief that you are the only one. The evidence is that you are not.

A few things I want to say directly, within the limits of what a counsellor can responsibly offer.

Loneliness is not a personal failing. The research in this report shows loneliness is patterned by life stage and circumstance. If you are a new mother, a carer, a recent migrant, a woman with disability, a rural woman, a widow, or a young woman in transition, the numbers are already on your side. You are responding to something real.

One strong connection matters more than many weak ones. Australian research is clear that quality, not quantity, predicts felt loneliness. If you have one person who knows you well enough to notice when you are not yourself, that is a foundation.

Name it when you can. Ending Loneliness Together research shows that almost half of Australians associate loneliness with negative traits. That stigma makes many women hide it. Saying "I have been feeling lonely lately" to someone safe is a small act with real evidence behind it.

Small, repeated contact beats big one-offs. A weekly walk with the same person. A standing phone call with a sister. A volunteer shift. The interventions with the strongest Australian evidence are built on regular, low-intensity contact, not on grand gestures.

If loneliness is sitting alongside anxiety, low mood, or thoughts of self-harm, that is a signal to reach out for support. A counsellor can sit with you while you work through what is underneath the loneliness. A GP can help you understand whether clinical support is appropriate. Both can be useful. Both are your choice.

On what I cannot do. As a registered counsellor, I cannot diagnose, prescribe, or provide clinical treatment for a mental health condition. I can offer a safe, confidential space to work through loneliness, grief, identity, relationships, and the transitions that cluster behind it. If you need clinical assessment, a GP or psychologist is the right first step and I will say so.

---

8. Help-seeking strip

If you are struggling, please reach out. You do not need to be in crisis to call these services.

• Lifeline. 13 11 14. 24 hours.
• Beyond Blue. 1300 22 4636. 24 hours.
• PANDA (perinatal anxiety and depression). 1300 726 306. Monday to Saturday.
• 13YARN (Aboriginal and Torres Strait Islander crisis line). 13 92 76. 24 hours.
• 1800RESPECT (family, domestic, and sexual violence). 1800 737 732. 24 hours.
• QLife (LGBTQIA+ peer support). 1800 184 527. 3pm to midnight.

If you would like to speak with Aana about counselling support, the Discovery Call is free and bookable at saferefuge.com.au. Sessions are not Medicare-rebatable (this is standard for ACA-registered counsellors) and no GP referral is required.

---

9. Methodology and sources

Scope. This report is a synthesis of existing Australian public data. No primary data collection was conducted. Every statistic in the body of the report is cited to a specific source.

Inclusion criteria. Sources had to be: (a) Australian or Australian-specific; (b) published by a recognised research, government, or peak body; (c) available for public verification; (d) published between 2019 and 2026 (with older foundational work cited where current equivalents do not exist).

Limitations.

• Loneliness is measured differently across the sources cited. The HILDA UCLA scale, the ABS survey, and the Ending Loneliness Together instruments do not always agree on prevalence, which is why prevalence quoted in this report ranges from roughly 15 per cent (HILDA, narrower definition) to 32 per cent (State of the Nation, broader definition). Readers should compare within, not across, instruments.
• Some chart specs describe data that is publicly reported but not published in the exact chart form described. The chart specs are instructions for a designer, not assertions that the chart already exists.
• Intersectional data (for example, rural Aboriginal women, CALD women with disability) is thin in the Australian public evidence base. This report does not invent numbers to fill those gaps.

Key sources.

• Australian Institute of Health and Welfare. Australia's Welfare 2023: Data Insights, Chapter 2 (Social isolation, loneliness and wellbeing). aihw.gov.au
• Australian Institute of Health and Welfare. People with Disability in Australia. Social inclusion and community support section.
• Australian Institute of Health and Welfare. Indigenous Health Performance Framework, Measure 1.18 (Social and emotional wellbeing).
• Australian Bureau of Statistics. General Social Survey: Summary Results, Australia, 2020.
• Australian Bureau of Statistics. National Study of Mental Health and Wellbeing, 2020 to 2022.
• Australian Bureau of Statistics. 2021 Census (living arrangements).
• Melbourne Institute. HILDA Statistical Report (annual, waves 1 to 23).
• Ending Loneliness Together. State of the Nation Report: Social Connection in Australia 2023.
• Ending Loneliness Together. Why We Feel Lonely: A Deep Dive (2024). Lim, Smith, Owen, Qualter.
• Ending Loneliness Together. A Call for Connection: Understanding and Addressing Youth Loneliness in Australia (2025).
• NSW Parliament, Standing Committee on Social Issues. Inquiry into the Prevalence, Causes and Impacts of Loneliness in New South Wales, Report No. 65 (2025).
• Lim, M. et al. 'The prevalence of chronic and episodic loneliness and social isolation from a longitudinal survey.' Scientific Reports, 2023.
• Aitken, Z. et al. 'Disability-related inequalities in the prevalence of loneliness across the lifespan: trends from Australia, 2003 to 2020.' BMC Public Health, 2024.
• Nowland, R. et al. 'Experiencing loneliness in parenthood: a scoping review.' Perspectives in Public Health, 2021.
• PANDA. How common are mental health issues in the perinatal period?
• Welch, V. et al. 'In-person interventions to reduce social isolation and loneliness: an evidence and gap map.' Campbell Systematic Reviews, 2024.
• Australian Longitudinal Study on Women's Health analysis. Communications Psychology, 2025.
• National Mental Health Commission. The Mental Health and Social and Emotional Wellbeing of Aboriginal and Torres Strait Islander Peoples (2024).

---

Media summary

The Australian Women's Loneliness Report: what the data tells us. A synthesis of existing public evidence by Safe Refuge Counselling, a private online and in-person counselling practice for women based in Mount Barker, South Australia.

Australian public data is unambiguous on one point. Women are not uniformly more lonely than men, but they carry a specific loneliness burden that tracks with life stage and context. In 2023, the HILDA survey put female loneliness at 16 per cent versus 15 per cent for men, while the Ending Loneliness Together State of the Nation Report put women at 32 per cent and men at 31 per cent. Different instruments produce different headline numbers, but the pattern holds.

The burden is not evenly spread across women. Young women aged 15 to 24 show the sharpest rise in loneliness over the past decade. Women in the perinatal period, drawing on international scoping reviews and PANDA's Australian estimates, are at elevated risk of co-occurring loneliness and mental health difficulty. Mid-life women in the Australian Longitudinal Study on Women's Health show that sustained loneliness trajectories are associated with markedly lower quality of life. Women aged 75 and over are the life stage at highest overall prevalence, driven partly by 40 per cent of them living alone.

Context multiplies risk. Women with disability are nearly twice as likely to report loneliness (29 per cent versus 17 per cent). Rural loneliness sits around 35 per cent. Older LGBTQIA+ women face additional loneliness drivers linked to community connection and experienced discrimination. Aboriginal and Torres Strait Islander women's loneliness sits inside a broader social and emotional wellbeing framework in which cultural connection is foundational. CALD women face language, digital, and network barriers the general-population data do not fully capture.

Loneliness and mental health travel together. People who are often lonely have a pooled adjusted odds ratio of about 2.33 for new-onset depression. People with mental health conditions are 2.9 times more likely to experience persistent loneliness. The causal direction is bidirectional, and the clinical implication is that reducing loneliness is a legitimate mental health strategy.

The evidence on what works is uneven but real. Group-based social-identity interventions, social prescribing, community-connection programs, and befriending have the strongest emerging Australian evidence. Digital-only and campaign-only responses have not yet demonstrated sustained effect.

For women considering whether their own loneliness warrants support, the practical message from the data is simple. Loneliness is common, patterned, and responsive to intervention. It is not a character flaw. Small, repeated, safe contact is more useful than waiting for the perfect moment. And if loneliness is sitting alongside low mood, anxiety, or difficulty coping, speaking to a counsellor, GP, or one of the national helplines listed above is a reasonable next step.

Contact. Aana Carpenter, registered counsellor (ACA Level 1), Safe Refuge Counselling. saferefuge.com.au. Available for interview on the findings in this report.

---

Compliance check notes.

This report is a synthesis of existing public data. No claim in the body of the report goes beyond the cited source. The 'What to do if you are lonely' section stays within counsellor scope, distinguishes counselling from clinical treatment, and directs readers to a GP for clinical assessment. Mindframe safe-messaging principles applied throughout. No specific methods of self-harm named. Faith content not included (secular-audience report). Brand fully disambiguated in opening line and media summary. Australian English; no em dashes.