918 Girls for Every 1,000 Boys - Why the Birth Ratio Is Improving and Where It Still Lags

What does the new number say?

India now records 918 girls for every 1,000 boys at birth. That is the headline from the SDG National Indicator Framework Progress Report 2026, released by the Ministry of Statistics and Programme Implementation (MoSPI) on National Statistics Day, 29 June 2026. In the words of a summary of the report’s key findings, the sex ratio at birth improved from 896 to 918 females per 1,000 male live births between 2015-17 and 2022-24.

That is a gain of 22 points in about seven years. It means that, for every 1,000 boys born, 22 more girls are now being born and counted than at the start of the decade.

Is it good news? Yes, with two honest caveats. First, 918 is still below what biology predicts. Second, different sources give different numbers, and a reader who picks one without knowing why can easily be misled. This article lays out all of them side by side.

The first two rows come from the Sample Registration System (SRS) series, which pools births over three years. The NFHS rows come from a household survey. They measure the same thing in different ways, which is why they do not match.

What is the sex ratio at birth, and why does 952 matter?

The sex ratio at birth (SRB) counts girls born per 1,000 boys born. It is different from the overall sex ratio (all women per 1,000 men) and from the child sex ratio (girls aged 0 to 6 per 1,000 boys of that age). The birth figure is the cleanest test of whether sex-selection is happening, because it measures what happens before any later difference in care, nutrition or survival.

Biology sets a baseline. More boys than girls are born in every population, roughly 105 boys for every 100 girls. A CEDA analysis at Ashoka University puts the natural ratio at about 952 girls per 1,000 boys. A figure far below that, year after year, cannot be explained by chance. It points to sex-selective abortion, or to girls going missing from birth records.

So the useful question is not just whether India’s number is rising. It is how far India still sits from 952. At 918, the gap is about 34 girls per 1,000 boys. At NFHS-5’s 929, it is 23. Both are better than a decade ago, and neither is closed.

Why do SRS, NFHS-5 and registration data disagree?

Three data systems report on births in India. Each answers a slightly different question.

SRS: a continuous sample of registered events

The Sample Registration System, run by the Registrar General of India, tracks births and deaths in a sample of villages and urban blocks, with a check by an independent survey. Because the sample for any one year is small, results are pooled over three years (2015-17, 2017-19, 2022-24). That smooths noise, but it also means each number covers a window, not a single year. SRS is the series MoSPI uses for the SDG indicator.

NFHS: what mothers report in a household survey

The National Family Health Survey asks women about every birth in the last five years. The count depends on memory, on who is interviewed, and on whether a girl who died early is still reported. It reaches a large sample and covers every state, which is why NFHS is the standard source for state-level comparison. But a five-year window means it describes births that happened between roughly 2015 and 2020, not the year of the survey. NFHS-5 reports 929 for India, against 919 in NFHS-4.

Civil registration: fast, but incomplete

States such as Haryana publish monthly or half-yearly birth ratios from the Civil Registration System (CRS). These are timely and can show district level, but they only count births that are registered, and registration can lag by weeks. A dip in one month may be a late-filing effect rather than a change in behaviour. A civil surgeon quoted by The Wire named delayed registration as one reason a short-period figure fell.

How to read them together

• Use SRS for the national trend over time.
• Use NFHS to compare states and groups such as rural and urban, or first-born and later-born.
• Use CRS to track a district month by month, but wait for the year to close before drawing a conclusion.
• Never subtract a figure from one source from a figure from another. A 929 in NFHS-5 and a 917 in SRS are not a contradiction. They are different instruments.

All three tell the same broad story: the ratio has been moving up since the mid-2010s. The precise level depends on which one you use.

Where has the ratio improved, and where does it lag?

National averages hide large state differences. The most useful example is Haryana, the state with the weakest record for years.

Haryana: from the worst in the country to a visible turnaround

In the 2011 Census, Haryana had the lowest child sex ratio of any large state, at about 830 girls aged 0 to 6 per 1,000 boys in the provisional count (the final figure was 834), as set out in a Tribune report on the Census data. Haryana was the centre of the debate on missing girls.

Then the numbers moved. Civil registration data summarised by Drishti IAS show Haryana’s ratio rising from 871 in December 2014 to 922 in December 2020. NFHS-5 independently shows Haryana’s birth ratio rising from 836 to 893, a gain of 57 points. Two different systems pointing the same way is a strong signal that something real changed.

The gains were not a straight line. The ratio stood at 914 at the end of 2021 and 916 up to June 2022, according to The Tribune, then reached 917 at the end of 2022 before slipping to 906 in mid-2023, as reported by The Wire. That slide led state officials to talk of a stall, though part of it may be the registration lag described above.

Haryana at a glance: the numbers in one table

Read the table by source type, not as one line. The Census number counts children aged 0 to 6, while the others count births. The household survey and the registration series cover different groups of mothers. What matters is the direction: sharply up from 2014, then a plateau in the low 910s, with a dip that may be partly a registration effect.

Inside the state, the gap between districts is wide

In the first half of 2022, the Tribune reported that 11 Haryana districts sat below the state average. Faridabad (879) and Mahendragarh (881) were lowest. Fatehabad (987) and Jind (959) were far above. Fatehabad had climbed from the 20th place the year before. A state average of about 916 held districts ranging from 879 to 987.

That spread is the real lesson. When a district can move 88 points in a year, leadership, enforcement and local attention matter more than any state slogan. It also means a district that slips can drop quickly.

Other states

Punjab, Haryana’s neighbour, still stands low at 904 in NFHS-5, according to CEDA. Northern and western states have long had the weakest ratios, with some southern and eastern states closer to natural levels. The national figure rises only if the weakest states improve. For that reason a state-by-state reading, rather than a single headline, is the one that guides policy.

Why does the birth ratio matter beyond the statistics?

A skewed ratio is a measure of what families believed a daughter was worth. It is also a social forecast. A generation born with too few girls grows up into a marriage market with too few brides. Researchers who study northern India have long warned that this can raise the pressure on poorer men who find it hard to marry, and can feed the buying and trafficking of brides from other regions. Those effects take twenty years to show, which is why a fix that comes late is harder to see and harder to credit.

There is also a plain rights point. The Constitution promises equal worth to every citizen. A pregnancy that ends because the foetus is female is a decision made on that basis alone, and a state that tolerates it at scale is failing its own promise. That is the case for treating the birth ratio as a governance indicator, which is why it sits among the national SDG indicators in the first place.

Finally, the ratio is one of the few social measures that can improve quickly when enforcement and attention line up. Child malnutrition or literacy moves over decades. A district’s birth ratio can move within a year. That makes it a useful test of whether a local administration is paying attention.

What drove the improvement?

No single cause explains a change of this size. Three forces stand out, and each has limits.

The PCPNDT Act, finally enforced

The Pre-Conception and Pre-Natal Diagnostic Techniques Act, 1994, bans sex determination before birth and requires every ultrasound centre to register. The law was weakly enforced for years. Convictions were rare, and many clinics ignored the register. The pressure changed in the 2010s, when courts, health departments and civil-society groups pushed for decoy operations, tracking of ultrasound machines and action against doctors. Enforcement is the part of the story with the strongest link to the Haryana recovery, because it is the one tool that directly stops the act of sex selection.

A government statement reported by IANS credited PC-PNDT enforcement for part of the national improvement in the 2021-23 reading of 917.

Beti Bachao Beti Padhao: a message, not a mechanism

The scheme was launched in January 2015 from Panipat in Haryana, aimed first at districts with low child sex ratios. It paired a communication campaign with a push on enforcement and girls’ schooling. Its strength is that it made son preference a topic for public conversation, and gave local officials a mandate. Its weakness is that a campaign does not change a family’s economics by itself. A household decides on the basis of dowry, old-age security and inheritance, and a slogan does not reach those directly. The scheme is best read as a spur to other action, not as the cause by itself.

Girls’ schooling and the changing value of a daughter

The slower force is girls in school. When a girl stays in school, marries later and earns, the cost-benefit logic that favours sons weakens. Families that see a daughter as a future earner, and as a person who will look after ageing parents, behave differently. This links to our reading of women’s labour force participation, where the same limit shows up: India educates more girls than ever but has not yet given them equal chances to work. Families notice that gap too. The same pattern shows in our look at the gender pay gap in India and in the numbers on women in science and engineering, where degrees do not yet turn into equal careers.

What the evidence does not show

It would be tidy to say these three policies added up to a 22-point gain, and leave it there. The evidence does not allow that. Data on birth ratios cannot separate the effect of the law from the effect of the campaign and from long-run social change. Researchers who study sex selection also point out that fertility has fallen, and small families make the pressure for a son stronger, not weaker. That makes the gains more fragile than they look.

What is still missing?

• The gap to 952 is open. 918 is not a natural ratio. Until it reaches the mid-940s, sex-selection is still happening somewhere.
• Unequal progress. The states and districts that most need change are the ones that moved least. A national average can rise while pockets stay low.
• Enforcement is uneven. Few people are convicted under PCPNDT compared with the number of complaints. Inspections vary by district, and sex determination has moved to informal, harder-to-trace channels.
• The demand side is untouched. Dowry, which remains common despite being illegal, and a son-first idea of old-age support, continue to drive the preference. No scheme here changes that as quickly as a law can ban a scan.
• Data is slow. The latest SRS reading of 2022-24 arrives in mid-2026. A district that slides today may not be visible for two years, except through CRS, which has its own lag.
• Missing girls after birth. The birth ratio does not capture neglect that comes later, in nutrition, health care and schooling. A birth count that looks healthy can coexist with higher child mortality for girls.

What to look for in the next data release

Three checks will tell a reader whether the gain is holding. First, whether the weakest states close the gap with the national average or fall further behind it. Second, whether SRS, NFHS and registration data all keep pointing upward, since one source alone can mislead. Third, whether the figure for first-born girls and for later-born girls is similar. In many families the first child is accepted whatever the sex, and the pressure shows up at the second or third birth, so a ratio that looks healthy overall can hide sharp skews for later births.

What works at the local level?

The Haryana story suggests five practices that district officials and citizens can copy.

1. Publish district-level data every month. Haryana’s open reporting made it possible to see who was improving. What is measured in public gets attention.
2. Hold the district collector accountable. The jump of Fatehabad and Jind shows the effect of an official who treats the ratio as a performance target.
3. Track every pregnancy. Registering pregnancies early, and following up on those that end without a recorded birth, closes the gap where sex-selective abortion hides.
4. Reward informers and protect them. Cases tend to be broken by people inside the system. Safe channels to report a clinic matter.
5. Work with families, not only clinics. Mothers-in-law, fathers and village elders influence the decision. Programmes that speak to them directly, and celebrate the birth of a daughter in public, change what neighbours expect.

What you can do

• Report sex determination. It is illegal for a clinic to tell the sex of a foetus and for a family to ask. Complaints can be made to the district appropriate authority (usually the Chief Medical Officer) under the PCPNDT Act, or through the state health helpline.
• Do not ask, and do not allow it to be asked. If someone in your family or circle asks a doctor, say plainly that it is a crime.
• Refuse and discourage dowry. The demand for it is one of the quiet drivers of son preference.
• Keep girls in school. Support a girl in your family or neighbourhood to finish Class 12 and beyond.
• Ask for local data. Ask your district administration or panchayat what the birth ratio in your area is, and whether it is published.
• Celebrate daughters. A small gesture, such as a naming ceremony, a plant or a school savings account for a newborn girl, changes how a community talks.

Frequently asked questions

Is India’s sex ratio at birth 918 or 929?

Both are correct, from different sources. 918 comes from the SRS-based MoSPI SDG report covering 2022-24. 929 comes from NFHS-5, a household survey covering births in the five years before 2019-21. They measure similar things at different times and by different methods.

What is the natural sex ratio at birth?

About 952 girls per 1,000 boys, according to the CEDA analysis. More boys than girls are born everywhere, so a ratio below 1,000 is normal. A ratio well below 952 over many years is a sign of sex selection.

Why is Haryana used as the example?

Because it had the lowest child sex ratio among large states in the 2011 Census, and then showed one of the clearest improvements across several data systems. It also shows how fragile progress can be.

Does Beti Bachao Beti Padhao deserve the credit?

It played a part, mainly by raising public awareness and focusing officials on low-ratio districts. Researchers cannot separate its effect from that of PCPNDT enforcement and wider social change, so no single scheme can claim the full gain.

Is sex determination illegal in India?

Yes. The PCPNDT Act, 1994, bans it. Both the person who asks and the clinic that tells can be prosecuted.

How often is the sex ratio at birth updated?

SRS reports it in three-year windows, and NFHS roughly every five years. State civil registration data can be read monthly, but short-period figures swing with registration delays.

Why does a ratio under 1,000 not mean something is wrong?

Because more boys than girls are born everywhere. The expected figure sits near 952 girls per 1,000 boys, not 1,000. The concern is a persistent figure well under that level, as India’s has been.

Can a family be punished for asking about the sex of the baby?

Yes. The PCPNDT Act holds both the person who seeks sex determination and the clinic that provides it responsible, and it requires scan centres to register and keep records.

The bottom line

918 girls for every 1,000 boys is real progress, backed by more than one data source, with Haryana as the clearest case. It is also incomplete. The natural level is near 952, the weakest districts still sit in the 870s and 880s, and a campaign cannot do the work of dowry reform and equal chances for women. The improvement shows that enforcement, public attention and schooling can shift a deep social habit within a decade. It does not show the habit is gone.