Is marriage right for you? Five separate questions to work through

“Is marriage worth it?” bundles several different questions together: whether a partner offers emotional support, shares the housework, contributes financially, or helps satisfy family expectations. Those things cannot be converted into a single unit and added up. This article separates them, using official statistics or meta-analyses where available and saying plainly where a calculation is not possible. It ends with a worksheet for your own circumstances, rather than a verdict on whether you should marry.

The registration figures, time-use surveys, and legal rules below concern mainland China. The research on health comes from several countries and is not a personalized forecast.

1. Separate the five questions

Money: the cash spent on customary marriage payments, the wedding, and housing, plus the legal ownership of assets. The source found no suitable official figures for the first group of costs; the ownership rules are set out in law.
Time: unpaid housework, caring for relatives, and making room for a partner’s career. The unit here is time, such as minutes per day.
Health: differences in mortality between married and unmarried groups. These establish associations, not how much marriage itself changes an individual’s health.
Legal exposure and the cost of leaving: property division, divorce procedures, and compensation for some unequal contributions.
Relationship quality: emotional support and the way conflict is handled. Studies can estimate associations, but these cannot be converted reliably into money or minutes.

The fifth question is hardest to put into numbers, yet often closest to what people actually want to know.

2. What the data can—and cannot—tell you

Marriage and divorce registration: two crude rates

China’s Ministry of Civil Affairs reported 6.106 million marriage registrations in 2024, down 20.5% from the previous year, with a crude marriage rate of 4.3 per thousand. There were 3.513 million divorces, with a crude divorce rate of 2.5 per thousand. Of these, 2.622 million were registered through civil affairs departments and 891,000 were granted or mediated by courts.

Both rates use the total population as their denominator: annual marriages or divorces divided by the average population, expressed per 1,000 people. “Crude” means that the population has not been broken into more specific groups. It tells you how many couples married or divorced that year per 1,000 residents. It does not tell you the probability that a particular marriage will eventually end. That requires following the same marriage cohort over time; the source did not find the necessary figures in an official original document.

Health: married groups have lower mortality, but this is not proof of cause

Three meta-analyses pooled findings across multiple studies. “All-cause mortality” means deaths from all causes combined.

Manzoli et al. (2007): 53 comparisons involving more than 250,000 older adults. Married people had lower all-cause mortality than a combined group of widowed, divorced or separated, and never-married people: RR 0.88, 95% CI 0.85–0.91, or about 12% lower relative risk. A risk ratio below 1 favors the married group in this comparison. Looking at unmarried categories separately, the reported ratios relative to married people were 1.11 for widowhood, 1.16 for divorce or separation, and 1.11 for never marrying—roughly 11%, 16%, and 11% higher. The pattern persisted across sex, study quality, and European versus American studies. A more conservative reanalysis gave RR 0.94, about 6% lower.
Roelfs et al. (2011): 95 publications covering more than 500 million people. Never-married people had higher all-cause mortality than married people: HR 1.24, 95% CI 1.19–1.30. This is about a 24% higher hazard, a measure of the relative rate of death during follow-up rather than a simple lifetime probability. The association was smaller at older ages and in more rigorous studies.
Wang et al. (2020): 21 studies covering roughly 7.89 million people. Unmarried status was associated with all-cause, cancer, and cardiovascular mortality in both sexes, with stronger associations for men for all-cause and cardiovascular mortality. The authors describe sex differences of 31% for stroke mortality and 9% for all-cause mortality among the never-married category. These figures come from comparisons of the marital-status relative risks between women and men; they are not a direct comparison of the absolute death rates of never-married men and women.

These were observational studies, not experiments assigning people to marriage. Healthier people, people with higher incomes, and people with stronger social networks may be both more likely to marry and more likely to live longer. That is a selection effect. Manzoli’s review also found signs of publication bias: more striking results may be more likely to appear in print, making a pooled result look more favorable.

Read the findings as “mortality was lower in the married groups studied,” not “getting married will cut your chance of death by 12%.” Much of the evidence comes from European and North American populations, so applying the size of these associations elsewhere needs caution.

Time: national gender gaps are context, not a prediction or ceiling

China’s third National Time Use Survey took place in May 2024. “Unpaid work” means work without pay. “Participants” means people who actually performed the activity on the survey day; the “participation rate” is the percentage who did so.

Among participants, unpaid work averaged 2 hours 45 minutes per day: 1 hour 52 minutes for men and 3 hours 29 minutes for women. Participation was 75.6% overall, 67.5% for men, and 83.9% for women.

Housework alone averaged 1 hour 17 minutes when spread over all residents. Among those who did housework that day, the average was 1 hour 59 minutes, 28 minutes less than in 2018, and participation was 64.9%. Companionship and care for family members averaged 30 minutes across all residents or 1 hour 46 minutes among participants, with participation of 28.4%.

The 2018 survey reported housework averaging 1 hour 26 minutes across all residents: 45 minutes for men and 2 hours 6 minutes for women. Participation was 40.4% for men and 75.6% for women. Everyday childcare averaged 36 minutes across all residents: 17 minutes for men, 53 minutes for women. Participation was 12.3% for men and 25.1% for women.

These publications break results down by sex, urban or rural residence, and age, but not by marital status. They therefore do not establish how many extra minutes of housework marriage creates. The 1 hour 37 minute gender gap among unpaid-work participants in 2024 is useful context for a discussion about sharing work. It is neither an upper limit on a couple’s imbalance nor a prediction of your own arrangement.

Money: default ownership, written agreements, and compensation for unpaid work

The source did not find suitable official totals for customary marriage payments, weddings, or marital housing, so it does not invent them. China’s Civil Code, effective January 1, 2021, does provide rules about property.

Article 1062: during marriage, wages, bonuses, remuneration for work, business and investment returns, intellectual-property income, and inherited or gifted property generally fall within jointly owned property. An inheritance or gift expressly designated for only one spouse is an exception. Both spouses have equal rights to deal with joint property.
Article 1063: separate property includes premarital assets, compensation for bodily injury, property a will or gift agreement specifies for one spouse alone, and personal daily-use items.
Article 1065: the couple can agree whether premarital and marital assets remain separate, become jointly owned, or follow a mixed arrangement. The agreement must be in writing and binds both spouses. Without an agreement, or where it is unclear, the default rules apply.
Article 1088: a spouse who has undertaken more responsibility for raising children, caring for older relatives, or supporting the other spouse’s work may seek compensation on divorce. The couple can agree the amount; otherwise the court decides.

These rules matter to both partners. The higher earner may want to focus on articles 1062 and 1065; the person doing more unpaid work on article 1088. A property agreement must be written, but these provisions do not make notarization a general requirement.

Leaving: the cooling-off period and court proceedings

Article 1076: if both spouses agree to divorce, they sign a written divorce agreement and apply in person to the marriage registration authority. They must have agreed on arrangements for children, property, and debts.
Article 1077: either spouse may withdraw the application during the 30 days after the authority receives it. After that cooling-off period, both must apply in person for the divorce certificate within the next 30 days. Failure to do so is treated as withdrawal.
Article 1079: when one spouse seeks divorce and the other does not agree, mediation may be attempted or a claim filed directly in court. The court also attempts mediation. If the relationship has irretrievably broken down and mediation fails, divorce should be granted. Listed grounds include bigamy or cohabitation with another person, domestic violence or abuse or abandonment, persistent gambling or drug misuse, and two years’ separation due to discord. If a court previously refused divorce and the couple then remains separated for a further year, a renewed claim should be granted.

Of the 3.513 million divorces in 2024, 891,000 went through court judgment or mediation. The registration cooling-off period applies to administrative registration, not court divorce proceedings.

Relationship quality deserves its own assessment

Robles et al. (2014) pooled 126 studies involving more than 72,000 people. Better marital quality was associated with better health. The effect size r expresses association: near zero means little relationship; a larger absolute value means a stronger one; a negative sign means the two measures move in opposite directions.

Reported associations were r = 0.07–0.21, including r = 0.11 for mortality-related outcomes and r = −0.13 for cardiovascular reactivity during conflict. These are small effects, but broadly comparable in scale to some associations between health behaviors, such as diet, and health. Some outcomes showed publication bias. Studies specifically testing sex differences generally found little evidence of a difference. Most of the evidence was observational, so it does not establish causation. Dhindsa et al. (2020) also discuss marital satisfaction and quality as relevant to cardiovascular outcomes.

The source did not find a meta-analysis directly comparing mortality in a poor-quality marriage with mortality among unmarried people. “A bad marriage is worse than staying single” therefore cannot be assigned a numerical estimate from this evidence. What can be said is that marital status and relationship quality are different questions. Looking only at the legal status leaves out a substantial part of the decision.

3. A worksheet for your own situation

Keep the categories separate. Do not add numbers with different units. If you cannot fill in a field, write “I don’t know.” That is useful information too.

What does each partner bring?

Money: income, debts, and premarital assets. Will you use the legal defaults or a written property agreement? If an agreement, what will it actually say?
Time: how much unpaid housework and care does each of you already do? For context, the 2024 participant averages were 1 hour 52 minutes for men and 3 hours 29 minutes for women. How will work be divided after marriage, after children, or when an older relative needs care? Whose career, if either, is expected to take a back seat?
Relationship quality: how often have you had conflict in the past six months, and how did each episode end? What did your partner do when you were ill, unemployed, or emotionally low? What did you do when the roles were reversed?
Emotional support: do you generally feel more resourced or more depleted after time together? This takes months of lived experience to judge; a list of credentials cannot answer it.

What would you contribute or give up?

Money: the actual costs of customary marriage payments, the wedding, and housing. What would that money otherwise be used for?
Time: extra daily housework and care, plus weekly time spent on family and social obligations.
Health: how would this relationship change your sleep, meals, and activity? That is closer to your circumstances than a pooled mortality estimate.
The cost of leaving: registration divorce involves the first 30-day cooling-off period and collection within the following 30 days. Court divorce follows article 1079. Could you manage the financial and practical demands of the process?

How can both partners protect their position?

Keep evidence of premarital ownership, such as purchase contracts, bank records, gift agreements, and wills. Does it establish that the asset existed before marriage or was intended for one spouse alone?
Decide whether to make a written property agreement, and whether it covers premarital property, later income, or both.
Discuss who will do more care and housework. Does that person understand article 1088’s compensation provision? Would you agree a method in advance?
Discuss debts: when could borrowing in one spouse’s name become joint debt, and does the other spouse know about it?

Put family pressure in a separate column

What do older relatives gain: reassurance, social approval, an end to pressure about marriage, or an expectation of future care?
Which of those are genuinely their benefits, and which are also yours?
You still bear your own financial, time, health, and legal costs. Benefits to relatives do not automatically become better health or a better relationship for you.
Once separated, you can decide how much weight this part deserves.

4. Common mistakes

Confusing the crude divorce rate with the divorce-to-marriage ratio. The 2024 crude divorce rate was 2.5 per thousand residents. Dividing that year’s divorces by that year’s marriages gives 3.513 / 6.106 ≈ 57.5%, but the numerator and denominator are different groups. Most people divorcing that year married in earlier years. If new marriages fall, the ratio rises even without a change in any marriage’s eventual likelihood of divorce. It does not mean “half of marriages end in divorce,” and it is not an indicator published in the cited Ministry bulletin.

Treating association as cause. The roughly 12% lower mortality estimate for married groups is a pooled association. Health, income, and social networks can influence both marriage and longevity. It should not replace an assessment of the relationship in front of you.

Treating marriage as a box to tick. Focusing on whether you are married can crowd out the question of how the relationship actually works. Relationship quality is a separate dimension.

Treating national averages as your future. A gender gap in a time-use survey is a population average. Default property rules can be changed by an appropriate written agreement. These figures provide context and a starting point for discussion, not a personal verdict.

Filling gaps with unsupported numbers. The source did not identify suitable official figures for customary marriage payments, wedding costs, or the cost of raising a child. Your own real budget is more useful than inserting an unrelated estimate simply to fill the blank.

Sources

Ministry of Civil Affairs (2025). 2024 Statistical Bulletin on the Development of Civil Affairs, section V(2)1 and note 5. https://www.mca.gov.cn/n1288/n1294/n1554/c1662004999980006190/content.html (PDF).
National Bureau of Statistics (2024). Third National Time Use Survey bulletins 1–3: https://www.stats.gov.cn/sj/zxfb/202410/t20241031_1957217.html, https://www.stats.gov.cn/sj/zxfb/202410/t20241031_1957216.html, https://www.stats.gov.cn/sj/zxfb/202410/t20241031_1957215.html; accompanying official Q&A: https://www.stats.gov.cn/sj/sjjd/202410/t20241031_1957218.html.
National Bureau of Statistics (2019). 2018 National Time Use Survey Bulletin. https://www.stats.gov.cn/sj/zxfb/202302/t20230203_1900224.html
National People’s Congress (2020). Civil Code of the People’s Republic of China, articles 1062, 1063, 1065, 1076, 1077, 1079, and 1088. National laws database; full text in the Supreme People’s Court Gazette.
Ministry of Civil Affairs (2020). Notice implementing Civil Code provisions on marriage registration, 民发〔2020〕116号. https://www.gov.cn/zhengce/zhengceku/2020-12/04/content_5567010.htm
Manzoli L, Villari P, Pirone GM, Boccia A (2007). Marital status and mortality in the elderly: a systematic review and meta-analysis. Soc Sci Med 64:77–94. https://doi.org/10.1016/j.socscimed.2006.08.031
Roelfs DJ, Shor E, Kalish R, Yogev T (2011). The rising relative risk of mortality for singles: meta-analysis and meta-regression. Am J Epidemiol 174(4):379–389. https://doi.org/10.1093/aje/kwr111
Wang Y, Jiao Y, Nie J, et al. (2020). Sex differences in the association between marital status and the risk of cardiovascular, cancer, and all-cause mortality: a systematic review and meta-analysis of 7,881,040 individuals. Glob Health Res Policy 5:4. https://doi.org/10.1186/s41256-020-00133-8
Robles TF, Slatcher RB, Trombello JM, McGinn MM (2014). Marital quality and health: a meta-analytic review. Psychol Bull 140(1):140–187. https://doi.org/10.1037/a0031859
Dhindsa DS, Khambhati J, Schultz WM, Tahhan AS, Quyyumi AA (2020). Marital status and outcomes in patients with cardiovascular disease. Trends Cardiovasc Med 30(4):215–220. https://doi.org/10.1016/j.tcm.2019.05.012