From Rhetoric to Recognition
Gender Roles and Fertility Decline in Turkey — 56th EBES Conference, Istanbul, 2–4 July 2026
1 Motivation: a fertility collapse with macroeconomic stakes
Turkey’s total fertility rate (TFR) fell to an all-time low of 1.48 in 2024 — well below the replacement level of 2.10 and below the OECD average. This is not a gradual drift: TFR has dropped by roughly one child in two decades.
| Indicator | Value | Note |
|---|---|---|
| TFR, Türkiye 2024 | 1.48 | All-time low; 2.10 = replacement |
| Change since 2000 | 2.6 → 1.48 | ~one child in two decades |
| Pronatalist agenda | since 2008 | “≥3 children” calls — little measurable effect |
Why should economists care? Sustained sub-replacement fertility drives population ageing, a shrinking labor supply, and pension-system strain — a first-order macroeconomic problem.
The puzzle is that fertility keeps falling despite an active pronatalist agenda promoted since 2008, which has had little measurable effect on outcomes (Attar 2024). This paper’s central claim is that the binding constraint is not household income — it is the unequal division of unpaid work inside the household.
2 The question and the argument
Research question. Why does fertility keep falling in Turkey even as the state turns pronatalist — and what does this reveal about the limits of policy?
When women’s roles expand in the public sphere but expectations in the private sphere stay rigid, fertility decline becomes self-sustaining — and cash incentives cannot fix a constraint that lives inside the household.
3 Theoretical framework
3.1 From the quantity–quality tradeoff to a fertility puzzle
Classical growth and fertility theory predicted that fertility would stabilise near replacement as incomes rose:
- Galor’s Unified Growth Theory (Galor 2011): development raises the returns to education, so parents trade child quantity for quality, lowering fertility toward replacement.
- Becker & Lewis (Becker and Lewis 1973): the quantity–quality tradeoff — rising incomes raise the “full” price of each child.
But fertility did not stabilise. Many advanced and emerging economies fell into “lowest-low” territory (TFR < 1.3). Crucially, the historic negative relationship between female labour-force participation and fertility reversed across the OECD (Ahn and Mira 2002; Feyrer, Sacerdote, and Stern 2008).
This raises the puzzle at the heart of the paper:
Why does fertility collapse in some countries but not others, under similar income and female-education levels? In 2020, Sweden and South Korea both had female participation above 70% — yet Sweden’s TFR (1.66) was nearly double Korea’s (0.84). Income and preferences alone cannot explain the divergence.
3.2 Goldin (2024): fertility as a credibility problem
The paper’s main conceptual basis is Claudia Goldin’s Babies and the Macroeconomy (Goldin 2024), which argues that the speed and sequence of modernization drive persistent fertility decline. A birth occurs only when a woman expects credible support — from a partner and from institutions — to combine children with a high-trajectory career.
Formally, a woman becomes a mother if and only if the expected value of childbearing exceeds its opportunity cost:
\[ p \cdot (K + \bar{w}) + (1-p)\cdot(K + w) \;\ge\; \bar{w}, \]
which rearranges to the mismatch condition:
\[ p \cdot (\bar{w} - w) \;\le\; K . \]
where
- \(p\) = the probability of credible support (a partner who shares, institutions that deliver);
- \(K\) = the value the woman places on children;
- \(\bar{w} - w\) = the career penalty for the flexibility that childrearing requires.
Under “greedy jobs” (Goldin 2014), flexibility carries a steep wage penalty, so women cannot simply buy their way out. Household bargaining and social norms (Chiappori 1992; Bertrand, Kamenica, and Pan 2015), together with persistent time-use gaps (Bianchi et al. 2012), lower the effective value of \(p\). When \(p\) is low or the penalty is high, rational childlessness follows even when \(K\) is high.
3.3 Where credibility comes from — and why generosity ≠ credibility
A broad literature identifies three domains that determine whether support is credible:
Labour-market rigidities. Greedy jobs and the motherhood penalty (Goldin 2014; Guner, Kaya, and Sánchez-Marcos 2020), plus job insecurity and income volatility, depress births (Adserà 2004; Sommer 2016; Lopes 2020).
Household constraints. Intensive parenting raises the time cost of children, falling mostly on mothers (Doepke and Zilibotti 2019); conversely, paternal involvement and paid paternity leave raise fertility and gender equality (Doepke and Kindermann 2019).
Welfare-state design. Spending and leave help (Gauthier 2007; Thévenon 2011), but design beats generosity: access and reliability are decisive, and poorly-designed (long, non-transferable) leave can backfire (Del Boca, Pasqua, and Pronzato 2009; Olivetti and Petrongolo 2017).
Turkey’s family-centered welfare regime places the burden of care on the household (Buğra and Keyder 2005; Grütjen 2008), and intergenerational norms sustain traditional caregiving roles even among highly educated women (Kavas and Thornton 2013; Kavas and Jong 2020). Meanwhile ideal family size remains high but intended fertility has dropped sharply, raising the prospect of a “one-child norm” (Abbasoğlu Özgören and Türkyılmaz 2023; Keskin and Çavlin 2023; Greulich, Dasré, and Inan 2016).
4 Evidence
4.1 Figure 1 — The scissors: participation up, fertility down
Female LFPR rose from roughly 25% in 2000 to over 47% by 2023, while TFR fell from 2.6 to 1.48 over the same period. The two series cross around 2011–2013, moving in opposite directions within a single generation. This is descriptive co-movement — not identified causation — but it is exactly the visual signature the mismatch framework predicts: as women enter public life without adequate work–care support, fertility falls.
4.2 Figure 2 — Cross-country: the unpaid-labour gap predicts fertility
Placing Turkey in the OECD reveals a clear negative relationship: fitted line TFR = 1.96 − 0.0032 × gap, R² = 0.33 — about a third of the cross-country variation in fertility tracks the unpaid-labour gap. Countries where unpaid work is more evenly distributed (Sweden, France, New Zealand) sustain TFRs around 1.7–2.1. Türkiye is the far-right outlier — the largest gap (~235 minutes/day) and one of the lowest fertility rates, alongside Japan and South Korea.
4.3 The mechanism in Türkiye: public advancement, private stagnation
| Indicator | Value | Source |
|---|---|---|
| Women’s weekly unpaid domestic & care work | ≈ 30 hours | Şahanoğulları et al. (2024) |
| Relative to men / OECD | 3× men; 2× OECD gap | — |
| Economic value of unpaid care | 1.4–3.3% of GDP | Aran & Aktakke (2016) |
Turkish women gained public-sphere roles without shedding private-sphere ones — the classic “second shift”. In Goldin’s terms, support is not credible (low \(p\)), so the effective cost of a child stays high and births are delayed or forgone. This makes Turkey a striking “Group 2” case: strong modernization, weak domestic convergence — the same logic that produced ultra-low fertility in Japan, South Korea, and Southern Europe.
5 From rhetoric to recognition: the 2025 speech
At the 2025 Family and Culture-Art Symposium, President Erdoğan noted that TFR is 1.38 for employed women versus 1.72 for others, but stressed that employment status is not the underlying cause — since fertility has also fallen among non-working women. Women, he argued, are “increasingly left alone” in childrearing, and “most of the burden unfortunately falls on women’s shoulders.”
Though framed in traditional cultural and religious language, this was the first high-level acknowledgment that household gender asymmetry — not merely economic conditions or values — underlies Turkey’s fertility decline. The diagnosis has moved inside the home, and the President urged fathers to “share responsibilities more fairly.” But recognition in a speech is not reform in institutions — which is the paper’s pivot.
6 Policy: cash is not enough — the constraint is structural
Why current policy underdelivers
- Financial incentives and family-support programs target income, not the unequal division of care.
- They leave the domestic labour split untouched, so the child’s effective cost stays on women.
- The predictable result is a limited, short-lived fertility response (Attar 2024).
What structural reform requires
- Fair parental leave — including well-paid, father-reserved leave (Doepke and Kindermann 2019).
- Accessible, affordable childcare as a genuine entitlement.
- Societal re-valuation of domestic and care work as shared, not female, labour.
Only reforms that make support credible — for men to share and for institutions to deliver — raise \(p\) and relax the constraint driving fertility decisions.
7 Conclusion
- Turkey’s fertility decline is best read as a Goldin-type mismatch: fast public-sphere modernization layered on rigid private-sphere norms (Figure 1).
- The binding constraint is the unpaid-labour gap, and the cross-OECD pattern places Turkey as the extreme case (Figure 2).
- The 2025 speech marks a genuine discursive shift: the state now names the household burden. Recognition has arrived.
- But recognition is not reform. Cash cannot substitute for shared, credible care.
- Contribution: a macro-grounded, policy-relevant reading that moves the pronatalist debate from wallets to households.
Next step. Measuring “credible support” across the OECD — the Credible Support Index (CSI), which operationalizes Goldin’s parameter \(p\) for 32 countries.
Delivery (15 min). Motivation ~1.5 min · Question/thesis ~1 min · Theory (3 sub-sections) ~4 min · Figures 1–2 ~3.5 min · Mechanism ~1 min · Speech ~1.5 min · Policy ~1.5 min · Conclusion ~1 min.
Lead with the number, not the speech — economists discount discourse analysis; the speech is evidence of norm change, not the thesis. The formal equation (Section 3) is what signals rigour to the room, including Zimmermann (Editor-in-Chief, Journal of Population Economics, present at the Meet the Editors session).
Anticipated questions. (1) Data for the time-use gap? — TÜİK time-use survey + OECD, Goldin (2024) method, 2009–2019. (2) Isn’t it just income? — No: incentives exist and fertility still fell; the gap is in care hours (Figure 2). (3) Causal identification? — The figures are descriptive; the identified cross-national test is the CSI paper. (4) R² = 0.33? — Modest but sizeable for a single cross-country correlate. (5) Endogeneity of the speech? — Treated as evidence of norm change, not a cause.