Demographic Documents
Income and employment characteristics associated with women’s first birth, 2017 to 2023

Overview of the study

This study examines the association between women’s income and employment characteristics and the likelihood of first birth. Specifically, it examines this association among women aged 22 to 39 who were childless, employed and living with a married spouse or common-law partner during the period 2017 to 2023.

Acknowledgments

The author gratefully acknowledges the many individuals who contributed to this article. Insightful comments on previous drafts were provided by Éric Caron Malenfant, Patrice Dion, Ana Fostik, France-Pascale Ménard, Jean-Dominique Morency, David Pelletier, Claudine Provencher, Paul Roberts, Daniël van Wijk, and Fei‑Ju Yang, which led to substantial improvements to the article. Jeffrey Fennelly, Mark Knarr and Nadine Laflamme assisted with preliminary data processing activities and Julie Faubert provided editorial assistance. Finally, Laurent Martel and Anne Milan provided indispensable support throughout the development of this project.

This research was supported in part by Women and Gender Equality Canada.

Highlights

• Women’s likelihood of first birth rises steadily with family income.
  From 2017 to 2023, among women aged 22 to 39 who were childless, employed and in a couple, those with higher family income were consistently more likely to have a first birth than those with lower family income.
• Income-related differences in the likelihood of first birth have widened over time.
  In 2017, women in the highest family income group were 2.1 times as likely to have their first child as those in the lowest-income group. By 2023, this ratio had widened to 2.5.
• Women's employment characteristics are associated with the likelihood of first birth.
  Regardless of income level, women employed in the public sector or in unionized jobs were more likely to have a first birth than other employed women.
• Family resources matter more than women’s individual earnings.
  While higher-earning women were more likely to have a first birth than lower-earning women, differences by family income were larger and widened more over time.
• COVID-19-era increases in first births were driven by women in higher‑income families.
  In 2021, the likelihood of first birth rose among women in the highest-income families but declined slightly among those in the lowest-income families.

Introduction and research questions

Canada’s total fertility rate (TFR) declined sharply over the last two decades, reaching a record low of 1.25 children per woman in 2024 (Statistics Canada, 2025). At the same time, childlessness is becoming more prevalent,^Note  and a substantial share of women who are currently childless intend to remain so.^Note  This decline in fertility has led to increased interest in the factors that influence the decision to have children, particularly the role of financial considerations amid economic uncertainty and the rising cost of living.

For much of the 20th century, fertility was negatively associated with income, especially women’s income (Doepke et al., 2022). Within traditional economic frameworks, this relationship was interpreted as reflecting the higher opportunity costs of childbearing for those with greater earning potential, particularly mothers who faced wage penalties when interrupting employment (Mincer and Polachek,1974; Becker, 1992).

However, evidence from the past several decades suggests that these associations are changing in high‑income countries. As high labour force participation and educational attainment among women have become the norm, opportunity‑cost explanations alone appear insufficient to account for observed fertility patterns (Goldstein, Sobotka and Jasilioniene, 2009; Myrskylä, Kohler and Billari, 2009). Recent international research indicates that the negative relationship between income and fertility has weakened or reversed in some high-income contexts, particularly when considering first births (Jalovaara and Miettinen, 2013; Hart, 2015; van Wijk, de Valk and Liefbroer, 2021). Also in these contexts, women’s stable employment and secure labour market attachment are increasingly associated with higher rather than lower fertility (Andersson and Scott, 2007; Kreyenfeld and Andersson, 2014; Miettinen and Jalovaara, 2020; Matysiak and Vignoli, 2026). These patterns may reflect broader institutional conditions that shape the compatibility of paid work and family life, influencing the extent to which employment and earnings facilitate, rather than hinder, entry into parenthood (McDonald, 2006; Esping-Andersen and Billari, 2015).

Alongside the reversal in the relationship between economic characteristics and fertility, recent studies of high-income countries show that fertility differences by income are widening. In particular, lower-income women are becoming increasingly unlikely to enter parenthood (van Wijk and Billari, 2024; van Wijk, 2024; Brini, Guetto and Vignoli, 2025), suggesting that financial constraints are growing in importance as a barrier to family formation. In contemporary societies, financial barriers to parenthood may stem from increased perceived economic requirements, widening inequalities in real economic resources, or the disproportionate impacts of macroeconomic shocks on those with lower socioeconomic status, among other factors (Adserà and Menéndez, 2011; Comolli, 2017; Doepke et al., 2022; Matysiak and Vignoli, 2023).

Whether these evolving relationships between income, employment and fertility are present in Canada remains largely unknown. Research in the Canadian context has been limited, as addressing these questions requires data that simultaneously capture women’s demographic, economic, employment, and family characteristics over time. Contemporary data sources typically capture these dimensions separately, cross-sectionally, or may lack sufficient sample sizes to examine detailed subpopulations.

To address this information gap, this study uses linked census and administrative data to examine how economic characteristics—specifically, family income, individual income, within-couple income distribution and employment characteristics—are associated with the likelihood of first birth among childless women^Note  aged 22 to 39 who are employed and in a couple. Specifically, it asks: Does the likelihood of having a first child vary by women’s income and employment characteristics, and did these patterns change from 2017 to 2023?

Analytical design and file construction

This study estimates differences in the likelihood of a first birth according to various economic characteristics of interest. Births and time-varying covariates are observed in annual intervals using logistic regression applied to person-year data. Each observation represents one year of exposure to the risk of having a first birth, and women contribute one person-year for each year in which they meet the inclusion criteria (described below in the “Sample construction and description” section).

The analytical approach broadly follows van Wijk (2024), facilitating international comparison. As in that study, lagged (t–2 years) measures of women’s economic characteristics are linked to the predicted probability of first birth in year t. A two‑year lag structure was selected to ensure that economic and family characteristics were measured prior to conception, consistent with the typical timing of conception^Note  and gestation.^Note  While childbearing intentions cannot be observed directly with the data sources used here, a two-year lag captures women’s circumstances at a point when decisions about childbearing are plausibly made, rather than during pregnancy or after birth.

Four data sources were linked to construct an analytical file with repeated measures of women’s demographic and economic characteristics as well as information on births to those women. The initial cohort was first identified in the 25% sample of the 2016 Census of Population. Women aged 15 to 44 were assumed to be at risk of a first birth (also referred to as childless or parity zero) if they did not usually live with at least one of their own children.^Note  This census-based measure of fertility history is known to slightly overestimate true childlessness, as some women may not usually reside with their children^Note  and the census may not capture all coresident parent–child relationships in certain complex households^Note  (Morency, Caron Malenfant and Daignault, 2018). This limitation was partially addressed through the incorporation of additional information from subsequent data linkage.

As shown in Figure 1, records from three administrative data sources were linked to the childless women identified in the 2016 Census 25% sample data. First birth events between 2016 and 2023 were identified from the Canadian Vital Statistics – Birth database (CVSB) via the Mother-centric data linkage. Information from the Canadian Vital Statistics – Death database (CVSD) was used to remove women from the sample following death, if applicable. Finally, lagged (t-2 years) measures of women’s annual income, employment, family and residential characteristics were obtained from the T1 Family File (T1FF).

Data table for Figure 1

The chart summarizes the data sources and years used in the study. The 2016 Census of Population (25% sample) was used to identify the initial cohort of childless women. The outcome of interest—whether a woman has a first birth in year t—was identified using the Canadian Vital Statistics - Birth Database for 2016 to 2023. The Canadian Vital Statistics - Death Database for 2016 to 2023 provided censoring information, identifying women who die before having a first birth.

Demographic and economic covariates were obtained from the T1 Family File for 2015 to 2021, measured two years prior to the outcome (t−2).

Outcomes were observed from 2017 to 2023, while covariates were measured from 2015 to 2021.

Source: Galbraith, N. (2026). Income and employment characteristics associated with women’s first birth, 2017 to 2023. Demographic Documents, Statistics Canada catalogue no. 91F0015M2026004.

Following linkage, records were reviewed and edited to resolve inconsistencies across sources based on each woman’s full set of observations over the study period.^Note  The resulting dataset is a longitudinal person-year file linking women’s pre-conception demographic and economic circumstances to observed first birth events.

Sample construction and description

To ensure that all ages included in the analysis were consistently observed across the full study period, the sample in each year is restricted to women aged 22 to 39. The former age was the youngest one present in all years^Note  (2017 to 2023) and the latter age was determined to be the most reasonable cut-off for obtaining a reliable estimation of women’s parity based on household composition.^Note 

Given that all economic and demographic characteristics of interest were captured in tax records, only those women who were successfully linked to a tax record in year t–2 were included in the analytical sample in year t. Among women aged 22 to 39, the rate of successful linkage decreased with temporal distance from 2016 (the year in which eligible women were identified in the census), from 86.7% in 2017 to 70.4% in 2023. This temporal decline is consistent with probabilistic linkages which rely on imperfect identifiers rather than a universal population register (St-Denis, Djomgang and Prats, 2026). As the interval between the census and administrative records grows, discrepancies in identifying information—such as changes in residence, surname, reporting inconsistencies, and missing data—are more likely to accumulate, reducing the probability of successful matches. Linkage rates also declined with women’s age, but this gradient emerged only in the later years of the study (Chart 1). This pattern suggests that the processes reducing linkage quality over time disproportionately affected older women in the sample.^Note 

Data table for Chart 1

Data table for Chart 1
Table summary
The information is grouped by Age (appearing as row headers), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using proportion (percent) units of measure (appearing as column headers).

Age	2017	2018	2019	2020	2021	2022	2023
	proportion (percent)
Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
22	85.6	85.9	85.4	84.8	85.2	83.6	83.1
23	86.3	86.8	85.5	84.7	84.6	83.0	82.3
24	86.7	86.8	85.6	84.4	83.8	82.0	81.3
25	87.2	87.3	85.3	83.9	82.8	81.1	80.3
26	87.6	87.6	85.3	83.1	81.9	80.0	78.9
27	87.3	87.7	85.2	82.4	80.8	78.4	77.3
28	87.5	87.9	84.6	81.4	79.5	76.7	75.4
29	87.5	88.0	83.8	80.2	77.4	75.3	73.2
30	87.2	88.3	83.4	78.5	75.4	72.4	71.0
31	86.6	87.9	82.6	77.0	72.9	69.8	67.8
32	86.6	87.4	82.1	75.3	70.5	66.7	64.4
33	86.7	87.6	81.0	73.7	68.3	64.1	60.9
34	86.4	87.9	80.9	72.8	66.8	61.8	58.2
35	86.0	87.4	81.3	72.3	65.2	60.3	56.0
36	85.2	86.8	81.1	73.2	65.5	59.4	54.8
37	85.2	85.9	81.5	74.2	66.8	59.9	54.6
38	85.8	85.8	81.0	74.9	68.0	61.6	55.5
39	85.8	86.6	81.9	75.7	69.4	63.4	57.6

The sample was further restricted to women who were employed and in a couple in year t-2 to facilitate a consistent comparison of economic conditions among those at risk of first birth, reflecting the joint and resource-dependent nature of childbearing decisions. As summarized in Table 1, women with employment income in year t-2 consistently exhibited a higher likelihood of first birth than women with income exclusively from other sources or those with no income. Additionally, women who were not living with a married spouse or common-law partner in year t-2 had distinctly lower likelihood of first birth in year t than those who were, consistent with the fundamentally different context for uncoupled women around childbearing decision‑making processes.

Several additional restrictions were applied to improve comparability across years and ensure consistent exposure to risk. As childlessness was defined as of Census Day (May 10, 2016), women who experienced a first birth later in 2016 were excluded to ensure comparable exposure intervals across all years. Additionally, women residing in Yukon in year t – 2 were excluded as births records registered in the territory between 2017 and 2022 were unavailable at the time of the data linkage.

A small number of women had tax records indicating their (self-reported) place of residence in year t – 2 was outside Canada; these women were excluded from the sample. However, it is expected that this adjustment underestimates the true level of emigration among the sample during the study period, as emigration was difficult to identify comprehensively with tax records.^Note 

Lastly, further exclusions addressed remaining misclassification of childlessness, where post-2016 administrative records indicated that a woman had children prior to an observed birth. This adjustment occurred in two cases: firstly, if a woman’s first birth event observed post-2016 indicated a previous birth (based on parity information on the birth record), and secondly, if family status information on tax records indicated she was already a parent.^Note 

Following these adjustments, the sample is a dynamic (time-varying) group of women drawn from those observed in the 2016 Census and linked administrative data. In each study year t, women were included (contributed one person-year) only if they met all of the following criteria: they were aged 22 to 39, had not yet given birth or died, had valid tax records in year t–2, were living with a married spouse or common-law partner in year t–2, and reported employment income in year t–2. In turn, a given woman could enter, exit and then re-enter the sample from year to year as their age, family, employment or tax record availability situation changed. In total, 958,407 person-years of exposure to first-birth risk were observed between 2017 and 2023, during which 113,414 women^Note  experienced a first birth, a rate of 118.3 births per 1,000 women at risk of first birth.

As shown in Chart 2, the sample became older over the study period. The total size of the sample ranged from a high of 140,576 women in 2018 to a low of 131,607 women in 2022. These year-to-year shifts in the size and composition of the sample reflect both the removal of women after they have a first birth and changes in the age profile of women meeting the inclusion criteria. In particular, the decline in the number of women at younger ages and their growing representation at older ages from 2017 to 2023 indicate that entry into co-residential unions and paid employment—and therefore into the sample—occurred at later ages over time. As a result, fewer women enter the sample at younger ages, shifting the age distribution to the right.

Data table for Chart 2

Data table for Chart 2
Table summary
The information is grouped by Age (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using number of women units of measure (appearing as column headers).

Age	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	number of women
Note: In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t-2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
22	3,659	3,624	3,405	3,315	3,319	3,069	2,932
23	4,724	4,858	4,610	4,245	4,346	3,835	4,017
24	6,312	5,908	6,105	5,600	5,580	5,075	4,919
25	8,298	7,702	7,133	7,197	7,132	6,411	6,464
26	9,945	9,520	8,959	8,138	8,688	7,977	7,876
27	10,993	11,145	10,568	9,795	9,479	9,270	9,264
28	11,361	11,722	11,594	10,918	10,527	9,815	10,374
29	11,376	11,608	11,786	11,419	11,275	10,544	10,455
30	10,778	11,050	11,157	11,133	11,418	10,609	10,572
31	10,214	10,087	10,359	10,320	10,827	10,262	10,381
32	9,179	9,369	9,215	9,368	9,622	9,709	9,735
33	8,105	8,418	8,304	8,210	8,499	8,503	8,935
34	7,187	7,469	7,556	7,552	7,456	7,499	7,856
35	6,500	6,677	6,753	6,946	6,883	6,632	6,922
36	5,887	6,023	6,138	6,221	6,393	6,172	6,194
37	5,216	5,495	5,664	5,702	5,823	5,787	5,800
38	4,971	5,046	5,222	5,231	5,407	5,350	5,502
39	4,715	4,855	4,867	5,061	5,028	5,088	5,138

Age-specific first-birth rates generally declined between 2017 and 2023, with the exception of 2021, when they increased across all ages (Chart 3). The absence of an apparent postponement trend should be interpreted with caution, as the sample is restricted to women who were both co-residing with a married spouse or common-law partner as well as employed—statuses that are themselves increasingly postponed to older ages over time.

Data table for Chart 3

Data table for Chart 3
Table summary
The information is grouped by Age (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using number of births per 1,000 women at risk of first birth units of measure (appearing as column headers).

Age	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	number of births per 1,000 women at risk of first birth
Note: Age-specific values have been smoothed through application of a 3-year moving average. In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t-2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
22	70	63	63	58	58	55	56
23	79	72	73	66	67	62	60
24	94	83	83	78	81	72	68
25	104	98	93	89	94	82	76
26	117	111	104	101	107	93	87
27	129	126	119	113	119	106	98
28	146	139	137	126	134	119	112
29	160	149	149	139	150	129	125
30	170	158	157	150	160	139	134
31	173	164	157	155	163	146	136
32	170	161	154	154	157	145	132
33	158	154	145	146	151	138	126
34	144	139	134	132	138	124	117
35	126	127	119	116	124	109	105
36	107	107	104	97	104	92	90
37	88	89	84	79	84	76	72
38	67	67	66	63	65	61	58
39	55	54	53	54	53	52	47

Other key demographic and economic characteristics of the sample are summarized in Tables 2 and 3.

Modeling strategy and specification of variables

Year‑specific logistic regression models were estimated to assess the association between women’s income and employment characteristics and the likelihood of a first birth. Logistic regression was selected given the discrete‑time structure of the data and the binary outcome of interest, that is, whether the woman had her first child within the year. Predicted probabilities were then derived post-estimation to facilitate comparisons across ages, years, and other subgroups, with associated measures of uncertainty (standard errors and confidence intervals) based on robust (Huber–White) variance estimation.^Note 

All models control for women’s age, province of residence in year t–2 and type of couple in year t–2. Specifically:

• Age was controlled flexibly using a cubic spline to account for the strong and non‑linear association between age and fertility. The spline allows the relationship between age and first-birth risk to vary smoothly with age in a non-linear manner.^Note 
• Province or territory of residence in year t–2 is included to account for differences in policy environments and other unobserved contextual factors specific to each jurisdiction that may influence childbearing decisions.^Note  Appendix 1 summarizes variations in first-birth likelihood by women’s province or territory of residence in year t-2.
• Although all women in the sample were in a couple, the type of couple (married or common-law) in year t-2 was controlled for as it was consistently associated with large differences in first-birth likelihood. This was the case even in Quebec where common-law couples are more prominent. For more details on Quebec-specific findings, see Appendix 2.

Family income, individual income, within-couple income distribution and employment characteristics—all measured in year t-2—were the economic characteristics of interest examined in separate models. Specifically:

• Women’s family income and individual income were analyzed separately. In both cases, after-tax income quintiles (excluding capital gains or losses) were defined within the (at-risk) sample by age and year to reflect women’s relative income position among peers, that is, relative to other childless, coupled and employed women of the same age. The use of income quintiles also facilitates comparisons across age and time and allows for non‑linear income gradients.
• To capture the organization of economic resources within couples, women’s share of total family income was included as a categorical variable (less than 25%, 25 to 49%, 50 to 74%, and 75% or more), distinguishing situations in which the division of family earnings was highly concentrated or more balanced.
• Three employment characteristics were modeled: industry, labour union membership, and self-employment status. Industry^Note  was grouped into four categories capturing broad differences in job stability, scheduling arrangements, access to employer-provided benefits and income predictability that may influence childbearing decisions: administrative, support or consumer-facing services; public administration, education or health; private professional, finance or knowledge services; logistics or production of goods; and unreported industry. Labour union membership^Note  serves mainly as a proxy for job protections, while self-employment reflects a distinct combination of greater scheduling flexibility and higher income volatility.

Additional sociodemographic characteristics reported in the 2016 Census—such as country of birth, generation status, ethnic or cultural origin and Indigenous identity—were assessed but excluded from the final models due to their negligible impact on the estimates related to economic characteristics and their volatility in disaggregated models. While educational attainment is known to correlate with fertility (Jalovaara et al., 2019, Vasireddy et al., 2023), its measurement in the census was insufficient given that for much of the sample, this status was collected during late adolescence or early adulthood, when schooling is often ongoing and final attainment remains undetermined, potentially leading to misclassification.

Census sampling weights were not applied, as they reflect the original survey design and do not account for additional selection introduced through linkage and analytic restrictions. As a result, women living on Indian reserves, in remote or northern areas, and in other hard-to-reach locations are overrepresented in the sample. This is because these areas were fully enumerated in the long-form census using canvasser interviews, whereas in other areas only a subsample of dwellings was included (Statistics Canada, 2017).

Table 4 summarizes the design of each logistic regression model and its correspondence to the study’s research questions.

Results

Given the large sample sizes involved, estimates are associated with relatively small standard errors, and observed differences can be interpreted with a high degree of precision. The analysis therefore emphasizes the magnitude and substantive interpretation of differences across groups. For clarity of presentation, confidence intervals are not shown in tables and charts; however, all estimates incorporate robust variance estimation, and measures of uncertainty are available upon request.

Higher family income is associated with a greater likelihood of first birth

Across all years examined, the likelihood of first birth followed a clear family income gradient: women in the top family income quintile consistently had the highest predicted probabilities of a first birth, while those in the bottom quintile had the lowest (Chart 4). In 2023, women in the top family income quintile had a 14.1% probability of experiencing a first birth, more than double the probability among women in the bottom quintile (5.6%).

Across the family income distribution, predicted probabilities of first birth generally declined between 2017 and 2023. The main exception occurred in 2021, when predicted probabilities temporarily increased for women in all family income quintiles except the bottom quintile. This rebound was most pronounced among women in the top family income quintile: between 2020 and 2021, their predicted probability of first birth rose 8.8% (from 16.3% to 17.8%). In contrast, first‑birth probabilities among women in the bottom quintile declined by 2.0% over the same period, from 6.8% to 6.7%.

Data table for Chart 4

Data table for Chart 4
Table summary
The information is grouped by Family income quintile in year t-2 (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using predicted probability (in percent) units of measure (appearing as column headers).

Family income quintile in year t-2	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	predicted probability (in percent)
Note: Predicted probability refers to the estimated likelihood of women in a given category to experience a first birth in a given year, conditional on being at-risk at the beginning of the year and averaging over the observed distribution of all other covariates in the model. The logistic regression models from which the predicted probabilities were derived also included the following covariates: age (modeled using a cubic spline); type of couple (married or common-law) in year t-2; and province or territory of residence in year t-2. In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t-2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
1st quintile	8.5	7.6	7.2	6.8	6.7	6.0	5.6
2nd quintile	10.6	10.0	9.5	9.0	9.5	8.6	8.3
3rd quintile	13.0	12.3	12.7	11.5	12.2	11.3	10.7
4th quintile	15.2	15.2	14.8	14.5	15.1	13.5	13.0
5th quintile	18.1	17.7	16.6	16.3	17.8	15.6	14.1

The association between income and first-birth likelihood has grown stronger

From 2017 to 2023, relative differences in the likelihood of first birth increased between the extremes of the family income distribution (Chart 5), owing to steeper proportional declines among women in the bottom family income quintile relative to those in the top quintile. In 2017, women in the top family income quintile were 2.1 times (or 210%) more likely to have a first birth than women in the bottom quintile, holding other factors constant. This ratio increased to 2.7 in 2021, before easing slightly to 2.5 in 2023.

Data table for Chart 5

Data table for Chart 5
Table summary
The information is grouped by Family income quintile in year t-2 (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using ratio units of measure (appearing as column headers).

Family income quintile in year t-2	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	ratio
Note: Predicted probability refers to the estimated likelihood of women in a given category to experience a first birth in a given year, conditional on being at-risk at the beginning of the year and averaging over the observed distribution of all other covariates in the model. The logistic regression models from which the predicted probabilities were derived also included the following covariates: age (modeled using a cubic spline); type of couple (married or common-law) in year t-2; and province or territory of residence in year t-2. In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t–2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
2nd quintile	1.2	1.3	1.3	1.3	1.4	1.4	1.5
3rd quintile	1.5	1.6	1.8	1.7	1.8	1.9	1.9
4th quintile	1.8	2.0	2.0	2.1	2.3	2.2	2.3
5th quintile	2.1	2.3	2.3	2.4	2.7	2.6	2.5

In contrast, the relative gap in first-birth likelihood between women in the fourth and fifth family income quintiles diminished over the period, from 19.0% in 2017 to 8.8% in 2023.

Patterns with respect to women’s individual income were broadly similar to those for family income, but less pronounced. After holding other characteristics constant, women with higher personal incomes were more likely to have a first birth (Chart 6) and the relative gap in this regard between the highest-and lowest-income women widened over time (Chart 7). As with family income, this gap widened because women in the bottom individual income quintile saw larger proportional declines (-29.3%) in the predicted probability of first birth than those in the top income quintile (-21.5%).

Data table for Chart 6

Data table for Chart 6
Table summary
The information is grouped by Individual income quintile in year t-2 (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using predicted probability (in percent) units of measure (appearing as column headers).

Individual income quintile in year t-2	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	predicted probability (in percent)
Note: Predicted probability refers to the estimated likelihood of women in a given category to experience a first birth in a given year, conditional on being at-risk at the beginning of the year and averaging over the observed distribution of all other covariates in the model. The logistic regression models from which the predicted probabilities were derived also included the following covariates: age (modeled using a cubic spline); type of couple (married or common-law) in year t-2; and province or territory of residence in year t-2. In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t-2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
1st quintile	9.0	8.5	8.3	7.8	7.7	6.9	6.4
2nd quintile	11.2	10.1	10.3	9.5	9.9	9.1	8.7
3rd quintile	13.0	12.2	12.2	11.7	12.7	11.2	10.7
4th quintile	15.2	15.2	14.2	13.9	14.8	13.1	12.6
5th quintile	17.0	16.7	15.8	15.3	16.3	14.6	13.3

Data table for Chart 7

Data table for Chart 7
Table summary
The information is grouped by Individual income quintile in year t-2 (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using ratio units of measure (appearing as column headers).

Individual income quintile in year t-2	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	ratio
Note: Predicted probability refers to the estimated likelihood of women in a given category to experience a first birth in a given year, conditional on being at-risk at the beginning of the year and averaging over the observed distribution of all other covariates in the model. The logistic regression models from which the predicted probabilities were derived also included the following covariates: age (modeled using a cubic spline); type of couple (married or common-law) in year t-2; and province or territory of residence in year t-2. In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t-2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
2nd quintile	1.2	1.2	1.2	1.2	1.3	1.3	1.4
3rd quintile	1.4	1.4	1.5	1.5	1.6	1.6	1.7
4th quintile	1.7	1.8	1.7	1.8	1.9	1.9	2.0
5th quintile	1.9	2.0	1.9	2.0	2.1	2.1	2.1

Within‑couple income distribution has an inconsistent association with the likelihood of first birth

Relative to women contributing less than one-quarter of total family income, those with higher contributions tended to have higher predicted probabilities of first birth (Table 5). However, there was considerable fluctuation in both direction and magnitude of effects over time when broken down by family income quintile, particularly among women in the bottom and top family income quintiles.

Women employed in the public sector have a consistently higher likelihood of first birth

Among the employment characteristics examined, industry of employment showed the strongest and most stable relationship with women’s likelihood of a first birth. From 2017 to 2023, women employed in the public sector—that is, public administration, education or health—had consistently higher predicted probabilities of first birth than those in other industries (Chart 8). Differences among the remaining industries were negligible.

The relative gap between public sector workers and others also widened over time. For example, in 2017, women in the public sector had a 16.3% higher predicted probability of first birth than their counterparts employed in administrative, support or consumer-facing service industries. By 2023, this difference had grown to 34.3%. This widening gap occurred because women employed in non-public sector industries saw stronger proportional declines in first-birth likelihood over the study period than public sector workers. Models stratified by women’s family income quintile yielded similar patterns. The size of this difference did not vary systematically by income quintile and showed no evidence of time trend.

Data table for Chart 8

Data table for Chart 8
Table summary
The information is grouped by Main industry of employment in year t-2 (appearing as row headers), Year (t), 2017, 2018, 2019, 2020, 2021, 2022 and 2023, calculated using predicted probability (in percent) units of measure (appearing as column headers).

Main industry of employment in year t-2	Year (t)
	2017	2018	2019	2020	2021	2022	2023
	predicted probability (in percent)
Note: Predicted probability refers to the estimated likelihood of women in a given category to experience a first birth in a given year, conditional on being at-risk at the beginning of the year and averaging over the observed distribution of all other covariates in the model. The logistic regression models from which the predicted probabilities were derived also included the following covariates: age (modeled using a cubic spline); type of couple (married or common-law) in year t-2; province or territory of residence in year t-2, family after-tax income quintile in year t-2, membership in a labour union in year t-2 and self-employed status in year t-2. In each year, the sample is restricted to women who met the following inclusion criteria: they were captured in the 2016 Census of Population (25% sample); were aged 22 to 39, had not yet had a first birth or died, had valid tax records in year t-2, were living with a married spouse or common-law partner in year t-2 and reported employment income in year t-2. Source: Linkage of Census of Population, 2016 (3901), T1 Family File, Canadian Vital Statistics - Birth database (3231), and Canadian Vital Statistics - Death database (3233).
Administrative, support or consumer-facing services	12.3	11.6	11.0	10.4	11.2	9.9	8.8
Public administration, education or health	14.3	13.8	13.7	13.1	13.5	12.2	11.8
Private professional, finance or knowledge services	12.3	11.8	11.1	10.8	11.4	10.3	9.3
Logistics or production of goods	12.5	12.1	11.5	11.0	11.6	10.4	10.1
Industry not reported	12.4	11.7	11.5	10.9	11.9	9.7	9.2

Women in unionized jobs have a higher likelihood of first birth

Women who belonged to a labour union had a higher predicted probability of birth than other women in all years (Table 6). However, these differences by union membership status were small relative to those observed with respect to income quintile and industry of employment, ranging from 4.6% higher (2017 and 2023) to 8.4% higher likelihood (2018). Moreover, unlike results for family income quintile and industry of employment, there was no time trend observed over the study period with respect to the predicted probability of first birth by union membership. Additional models stratified by family income quintile showed no evidence of an interactive effect of union membership and first-birth likelihood by family income quintile

Women who were self-employed had a slightly higher predicted probability of first birth than other women in the global model (that is, when combining women of all family income quintiles). However, these patterns were less consistent when examining women in each family income quintile separately (Table 7). Stratified results showed considerable variation in both direction and magnitude, with some quintiles exhibiting values below one in certain years and others showing relatively large positive deviations.

Discussion

This study provides new evidence about the relationship between women’s economic characteristics and the likelihood of becoming a parent during a period of historically low fertility and rising childlessness in Canada. The results show a clear and widening gradient by income, with women in higher‑income families consistently more likely to have a first birth than those in lower‑income families. These findings align with recent evidence from other high-income countries (van Wijk, 2024; Brini, Guetto and Vignoli, 2025) and therefore situate Canada squarely within a broader international pattern in which parenthood is increasingly associated with economic advantage.

Income-related differences in first-birth likelihood may reflect either variation in fertility intentions across income groups or differences in the capacity to realize those intentions. Fertility intentions could not be directly observed in the present study; however, prior research suggests that desired family sizes vary little by socioeconomic status or economic conditions, while more pronounced differences have been found in the realization of those intentions (Sobotka and Beaujouan, 2014; Beaujouan and Berghammer, 2019; Behrman et al., 2026). In this context, the income-related differences in entry into parenthood observed here are consistent with the role of financial barriers related to economic insecurity and the perceived feasibility of family formation, as well as broader economic uncertainty, which have been shown to influence both the timing and likelihood of childbearing (Schneider and Hastings, 2015; Comolli, 2023).

In 2021, Canada recorded a rare year‑over‑year increase in its total fertility rate (TFR), the first such rise since 2008. Similar short‑lived fertility rebounds were observed across many high‑income countries and have been interpreted as temporary disruptions associated with the COVID‑19 pandemic (Provencher and Galbraith, 2024; Sobotka et al., 2024). Within Canada, prior research showed the increase was driven primarily by births to Canadian‑born, rather than foreign‑born, women (Provencher, 2025). The present analysis extends this evidence by showing that increases in first-birth likelihood that year were concentrated among women in the highest‑income families. This pattern aligns with international findings that fertility responses to periods of uncertainty are uneven and selective, reflecting disparities in economic security (Luppi, Arpino & Rosina, 2022; Comolli, 2023; Bloom, Kuhn & Prettner, 2024).

After taking into account the family’s overall income position, the division of income within the couple was associated with small and inconsistent differences in first‑birth likelihood. This finding underscores that in the contemporary Canadian context—where most couples are dual-earner^Note  and more than three-quarters of employed women work full-time^Note —access to sufficient and stable economic resources could be more consequential for entry into parenthood than the precise division of those resources within the couple.

Women’s employment characteristics were also strongly associated with the likelihood of first birth. Women employed in the public sector consistently exhibited higher likelihoods of first birth than women in other industries, as did those who belonged to a labour union. These patterns are well documented internationally and are commonly linked to employment attributes such as greater job security and stability (Adserà, 2011; Hsu, 2023) and more family‑friendly working conditions (Fluchtmann, van Veen and Adema, 2023; OECD, 2024). Moreover, the persistence of these industry differences across the income distribution suggests that women’s employment context is independently associated with entry into parenthood, net of women’s economic resources.

Unlike industry or union membership, self‑employment did not have a stable association with entry into parenthood. The absence of a coherent pattern by family income quintile suggests that the relationship between self‑employment and first‑birth likelihood is context‑dependent, potentially varying with macroeconomic conditions and the specific nature of women’s self‑employment. Furthermore, these findings suggest that while some employment characteristics may support family formation, job flexibility alone does not necessarily provide a stable pathway into parenthood in the absence of economic security and institutional protections.^Note 

Several limitations should be considered when interpreting these findings. First, the study did not include information about fertility intentions, relationship quality, health status (including infertility), or preferences regarding work–family balance, all of which may mediate the observed relationships between income, employment characteristics, and entry into parenthood.

Second, economic resources were measured using cross-sectional income and lacked information about longer-term income trajectories or wealth, which may also influence childbearing decisions. Moreover, income quintiles were defined within the sample rather than the broader population, capturing relative differences within this group but not necessarily individuals’ broader economic reference points and context.

Third, features of the data and sample design may affect representativeness. Higher census undercoverage among young adults (Statistics Canada, 2019), combined with the exclusion of recent immigrants and non-tax-filers—who are more likely to be young and have lower incomes (Robson and Schwartz, 2020; St-Denis, Djomgang and Prats, 2025)— may bias the sample toward more economically advantaged populations. Additional restrictions, including limiting the analysis to women aged 22 to 39 who were both in a couple and employed, may further reduce generalizability—especially given that income itself may influence the likelihood of being in a stable co-residential union (Ferraretto, Hiekel and Vitali, 2025). Finally, because sample inclusion depended on successful record linkage and linkage success is not random, there is additional potential for selection bias. Taken together, these features likely led to an underrepresentation of lower-income populations, suggesting that the overall role of income in first-birth likelihood is likely understated in this analysis.

Despite these limitations, the consistency of the findings with recent evidence from other high-income countries suggests that the observed patterns reflect broader structural changes in the relationship between economic resources and entry into parenthood rather than artifacts of the data.

Future research could extend this work by examining transitions to second and higher‑order births; analyzing income trajectories over longer periods prior to parenthood; and exploring geographic variation in greater detail—for example, by distinguishing urban and rural areas within provinces or accounting for differences in housing costs and childcare availability across major population centres. Extending the analysis to men’s transitions^Note  to parenthood and to women’s post-birth income and employment trajectories would also provide a more complete picture of economic and fertility dynamics in Canada.

Conclusion

This study found that from 2017 to 2023—a period of rapidly declining fertility in Canada— women’s likelihood of becoming a parent differed by income. Among coupled, employed women, those with higher family incomes were consistently more likely to have a first birth than those with lower family incomes. This income-related gap widened over the period, despite declining first-birth probabilities across the income distribution, because women in the lowest-income families experienced larger declines than their higher-income counterparts.

In addition to income, employment characteristics were linked to the likelihood of first birth. Women employed in jobs offering greater security, benefits and more predictable working conditions were more likely to have a first birth, regardless of income level.

Together, these findings indicate that economic resources and security played an important and growing role in women’s transition to parenthood in Canada in recent years. At the same time, first-birth probabilities declined over the study period across all income and employment groups. This pattern suggests that additional factors beyond the economic characteristics examined in this study also contributed to recent fertility trends.

From a methodological perspective, this study combined census, tax, and vital statistics data to produce novel micro-level evidence of the association between women’s economic circumstances and the likelihood of becoming a parent. These findings highlight the value of linked administrative data for advancing understanding of fertility in Canada.

Appendices

Appendix 1: Differences in first-birth likelihood by province of residence

Table A1 presents the predicted probability of first birth by women’s province of residence in year t-2. Women residing in Quebec consistently held the highest predicted probability of first birth from 2017 to 2023, despite declining from 18.1% in 2017 to 14.5% in 2023. In contrast, Newfoundland and Labrador, Nova Scotia, Alberta, and British Columbia generally recorded the lowest predicted probabilities throughout. The size of provincial differences remained stable over time: in 2023, the highest probability was 77.2% greater than the lowest, similar to the gap observed in 2017 (78.1%).

Provincial rankings of first-birth probabilities differ from those based on the total fertility rate (TFR). For example, although Saskatchewan had the highest TFR in 2023, it ranked sixth in estimated first-birth likelihood among women in this sample. This difference occurs because the two measures capture distinct aspects of fertility and are based on populations with different compositions. The TFR summarizes overall fertility across all ages and birth orders, whereas the estimates presented here measure entry into parenthood among childless, employed and coupled women aged 22 to 39 while controlling for several sociodemographic characteristics. Provincial variation may therefore reflect both differences in fertility behaviour and differences in the share of women who are childless, coupled, and employed. In addition, provinces may have relatively high TFRs due to earlier timing of childbearing, higher-order births, or higher shares of births to unpartnered or unemployed women, even when first-birth likelihood within this sample is comparatively low.

Appendix 2: Quebec-specific findings

Quebec represents a distinct institutional and social context within Canada, characterized by a unique family policy regime that includes provincial maternal and parental benefits, subsidized early childhood education and care, and family benefits. Assessing the combined impact of these policies on fertility behaviour remains challenging, given variation across studies in the policy mechanisms examined, methodological approaches, comparison groups, and outcome measures. At the same time, Quebec has long exhibited family-formation patterns that differ from those elsewhere in Canada, particularly the greater prevalence of common-law unions and childbearing within these unions.^Note 

Against this backdrop, it was hypothesized that the relationships between women’s socioeconomic position, union type and first‑birth likelihood would differ between Quebec and the rest of Canada. To examine this, separate logistic regression models were estimated for women residing in Quebec and for those living elsewhere in Canada.

As shown in Table A2, the predicted probability of first birth followed a consistent income gradient in both regions over the study period. For example, in 2023, Quebec women in the top family income quintile were 2.8 times more likely to have a first birth than their counterparts in the bottom quintile, while the same ratio was 2.5 among women living elsewhere in Canada. From 2017 to 2023, the relative size of the gap between the highest and lowest income families widened to a similar degree among women living in Quebec (+22.7%) and those living elsewhere in Canada (+21.9%). These results indicate that despite Quebec’s distinct policy environment, income‑related differences in the transition to parenthood are prominent in the province and have widened in recent years.

In both Quebec and the rest of Canada, married women had higher probabilities of first birth than those in common‑law unions in all years. However, these differences were consistently smaller in Quebec. Moreover, although the relative advantage associated with marriage diminished between 2017 and 2023 in both regions, the decrease was more pronounced in Quebec. In sum, while marriage continues to differentiate first‑birth likelihoods among coupled women across Canada, its role is weaker in Quebec. This finding is consistent with previous evidence showing that the close link between marriage and parenthood diminishes in contexts where non-marital unions are prevalent and widely accepted as a setting for childbearing (Le Bourdais and Lapierre-Adamcyk, 2004; Hiekel and Castro-Martin, 2014; Laplante and Fostik, 2015).

References

Adserà, A. (2011). Where are the babies? Labor market conditions and fertility in Europe. European Journal of Population, 27(1): 1-32.

Adserà, A. and Menéndez, A. (2011). Fertility changes in Latin America in periods of economic uncertainty. Population Studies, 65(1): 37-56.

American College of Obstetricians and Gynecologists. (2013). Definition of term pregnancy. Committee opinion Number 579, November 2013.

Andersson, G. and Scott, K. (2007). Childbearing dynamics of couples in a universalistic welfare state: The role of labor-market status, country of origin, and gender. Demographic Research, 17(30): 897-938.

Beaujouan, E. and Berghammer, C. (2019). The gap between lifetime fertility intentions and completed fertility in Europe and the United States: A cohort approach. Population Research and Policy Review, 38: 507-535.

Becker, G. S. (1992). Fertility and the economy. Journal of Population Economics, 5(3): 185-201.

Behrman, J.A., Marshall, E.A. and Percheski, C. (2026). Economic resources and parity among US women: A conjoint experiment on preferred family scenarios. Demographic Research, 54(34): 1095-1124.

Bloom, D. E., Kuhn, M. and Prettner, K. (2024). Fertility in high‑income countries: Trends, patterns, determinants, and consequences. Annual Review of Economics, 16: 159-184.

Brini, E., Guetto, R. and Vignoli, D. (2025). A research note on the increasing income prerequisites of parenthood. Country-specific or universal in Western Europe? Econometrics Working Papers Archive 2025_05, Universita' degli Studi di Firenze, Dipartimento di Statistica, Informatica, Applicazioni "G. Parenti".

Comolli, C. L. (2017). The fertility response to the Great Recession in Europe and the United States: Structural economic conditions and perceived economic uncertainty. Demographic Research, 36(51): 1549-1600.

Comolli, C.L. (2023). Social climate, uncertainty and fertility intentions: from the Great Recession to the COVID-19 crisis. European Journal of Population, 39(35).

Doepke, M., Hannusch, A., Kindermann, F. and Tertilt, M. (2022). The economics of fertility: A new era. NBER Working Paper No. 29948.

Esping‑Andersen, G. and Billari, F. C. (2015). Re‑theorizing family demographics. Population and Development Review, 41(1): 1-31.

Ferraretto, V., Hiekel, N. and Vitali, A. (2025). Non-cohabiting partners’ economic characteristics and the transition to living together in Germany: A couple-level perspective. European Journal of Population, 41(25).

Fluchtmann, J., van Veen, V. and Adema, W. (2023). Fertility, employment and family policy: A cross-country panel analysis. OECD Social, Employment and Migration Working Papers, No. 299, OECD Publishing, Paris.

Goldstein, J. R., Sobotka, T. and Jasilioniene, A. (2009). The end of “lowest‑low” fertility? Population and Development Review, 35(4): 663-699.

Hart, R. K. (2015). Earnings and first birth probability among Norwegian men and women, 1995–2010. Demographic Research, 33(38): 1067-1104.

Hiekel, N. and Castro-Martin, T. (2014). Grasping the diversity of cohabitation: Fertility intentions among cohabiters across Europe. Journal of Marriage and Family, 76(3): 489-505.

Hsu, C-H. (2023). How women’s employment instability affects birth transitions: the moderating role of family policies in 27 European countries. European Sociological Review, 39(6): 935-956.

Jalovaara, M., Neyer, G., Andersson, G., Dahlberg, J., Dommermuth, L., Fallesen, P. and Lappegård, T. (2019). Education, gender, and cohort fertility in the Nordic countries. European Journal of Population, 35(3): 563-586.

Jalovaara, M. and Miettinen, A. (2013). Does his paycheck also matter? The socioeconomic resources of co-residential partners and entry into parenthood in Finland. Demographic Research, 28(31): 881-916.

Kreyenfeld, M. and Andersson, G. (2014). Socioeconomic differences in the unemployment and fertility nexus: Evidence from Denmark and Germany. Advances in Life Course Research, 21: 59-73.

Laplante, B. and Fostik, A.L. (2015). Disentangling the Quebec fertility paradox: The recent evolution of fertility within marriage and consensual unions in Quebec and Ontario. Canadian Studies in Population, 42(1-2).

Le Bourdais, C. and Lapierre‑Adamcyk, É. (2004). Changes in conjugal life in Canada: Is cohabitation progressively replacing marriage? Journal of Marriage and Family, 66(4): 929-942.

Matysiak, A. and Vignoli, D. (2023). Family life courses, uncertain futures, and the changing world of work: State-of-the-art and prospects. European Journal of Population, 40(19).

Matysiak, A. and Vignoli, D. (2026). The end of an era: The vanishing negative effect of women's employment on fertility. Population and Development Review, 52(1): 220-235.

McDonald, P. (2006). Low fertility and the state: The efficacy of policy. Population and Development Review, 32(3): 485–510.

Mincer, J. and Polachek, S. (1974). Family investments in human capital: Earnings of women. Journal of Political Economy, 82(2): S76–S108.

Miettinen, A. and Jalovaara, M. (2020). Unemployment delays first birth but not for all: Life stage and educational differences in the effects of employment uncertainty on first births. Advances in Life Course Research, 43-100320.

Morency, J.‑D., Caron Malenfant, É. and Daignault, P. M. (2018). Fertility and Aboriginal people in Canada: An overview of trends at the turn of the 21st century. Aboriginal Policy Studies, 7(1): 34-61.

Myrskylä, M., Kohler, H.‑P. and Billari, F. C. (2009). Advances in development reverse fertility declines. Nature, 460(7256): 741-743.

OECD (2024). Society at a Glance 2024: OECD Social Indicators, OECD Publishing, Paris.

Provencher, C. and Galbraith, N. (2024). Fertility in Canada, 1921 to 2022. Demographic Documents. Statistics Canada catalogue no. 91F0015m2024001.

Provencher, C. (2025). The contribution of foreign-born mothers to Canadian births from 1997 to 2024. Demographic Documents. Statistics Canada catalogue no. 91F0015m2025004.

Robson, J. and Schwartz, S. (2020). Who doesn’t file a tax return? A portrait of non-filers. Canadian Public Policy, 46(3): 305-443.

Schneider, D. and Hastings, O.P. (2015). Socioeconomic variation in the effect of economic conditions on marriage and nonmarital fertility in the United States: Evidence from the great recession. Demography, 52(6): 1893-1915.

Shapiro, G.D., Bushnik, T., Sheppard, A.J., Kramer, M.S., Kaufman, J.S. and Yang, S. (2016). Missing paternal data and adverse birth outcomes in Canada. Health Reports, 27(12): 3-9. Statistics Canada catalogue no. 82-003-X.

Sobotka, T., and Beaujouan, É. (2014). Two is best? The persistence of a two-child family ideal in Europe. Population and Development Review, 40(3): 391-419.

Sobotka, T., Zeman, K., Jasilioniene, A., Winkler-Dworak, M., Brzozowska, Z., Alustiza-Galarza, A., Németh, L. and Jdanov, D. (2024). Pandemic roller-coaster? Birth trends in higher-income countries during the COVID-19 Pandemic. Population and Development Review, 50(S1): 23-58. 

Statistics Canada (2017). Chapter 9 – Sampling and weighting for the long form. Guide to the Census of Population, 2016. Statistics Canada catalogue number 98-304-X2016001.

Statistics Canada (2019). Coverage Technical Report, Census of Population, 2016. Statistics Canada catalogue number 98-303-X.

Statistics Canada (2025). The Daily — Fertility and baby names, 2024. September 24, 2025.

Statistics Canada (2026). The Daily — Fertility and intentions: Socioeconomic factors. January 26, 2026.

St‑Denis, X., Nono Djomgang, C. and Prats, N. (2025). La science des données administratives canadiennes : guide pratique pour la recherche en sciences sociales. Institut national de la recherche scientifique. Groupe d’étude en statistique sociale.

van Wijk, D. C., de Valk, H.A.G and Liefbroer, A.C. (2021). Temporary employment and family formation: An income or insecurity effect? European Journal of Population, 37(4): 641-658.

van Wijk, D. C. (2024). Higher incomes are increasingly associated with higher fertility: Evidence from the Netherlands, 2008–2022. Demographic Research, 51(26): 809-822.

van Wijk, D. C. and Billari, F. C. (2024). Fertility postponement, economic uncertainty, and the increasing income prerequisites of parenthood. Population and Development Review, 50(2): 287-322.

Vasireddy, S., Berrington, A., Kuang, B. and Kulu, H. (2023). Education and fertility: A review of recent research in Europe. Comparative Population Studies, 48.