Maternal Mortality

OWID presents work from many different people and organizations. When citing this entry, please also cite the original data source. This entry can be cited as:

Max Roser (2017) – ‘Maternal Mortality’. Published online at Retrieved from: [Online Resource]


# Empirical View

# Maternal Mortality in the Long Run

The chart below shows the decline of maternal mortality. Going back to the 19th century and looking at countries that have the best health today we see out of 100,000 births about 500 to 1,000 ended with the death of the mother. This means every 100th to 200th birth led to the mother’s death.

Since women gave birth much more often than today the death of the mother was a tragic but not uncommon tragedy. This changed over the last century and today most rich countries have a maternal mortality ratio of 10 deaths per 100,000 births – the countries with the lowest maternal mortality reached a level of around 1% of the death rate in the 19th century.

The countries that achieved the lowest maternal mortality rate are Finland, Greece, Iceland, and Poland. For each 100,000 deaths 3 mothers die.

The 100-fold decline of maternal mortality is due to the modern scientific understanding of the cause of maternal mortality and the adoption of practices which appear surprisingly simple in hindsight. The common reason for the mother to die was puerperal fever (or childbed fever) which was caused by unhygienic medical staff and medical equipment by which the mother’s genital tract is infected during childbirth.

It was the physician Ignaz Semmelweis who first noticed the link between hygiene and the survival of mothers in the middle of the 19th century. He urged his colleagues to wash their hands with chlorinated lime solutions but was ignored. The germ theory of disease was not yet known and therefore he could not explain why there should be a link between hygiene and the survival of women during childbirth. The rejection by the medical community of the time turned Semmelweis bitter and every conversation he had revolved around childbed fever. He was eventually committed to an mental asylum where he died a miserable death. He was never to see how right he was and never knew how many mother’s lives he saved!

After Semmelweis’ death, when Louis Pasteur developed the germ theory of diseases, the recommendations of Semmelweis were finally adopted and maternal mortality started to decrease. A procedure as simple as the doctor washing his hands meant that puerperal fever – a killer of thousands of mothers – declined sharply. We see the decline in Finland over the course of the 2nd half of the 19th century. In the 20th century the availability of antibiotics made it possible to treat cases of puerperal fever and the death of a mother is today fortunately very rare. As it is often the case we see that it is much harder for a pioneer to make advancements than for a country that catches up later. The decline of maternal mortality in Finland began in the middle of the 19th century and didn’t reach today’s low level more than a century later. Malaysia in contrast achieved this progress in only a few decades.

# The Maternal Mortality Rate around the world

The risk of women dying during child birth is declining around the world as the following visualisation shows. In this visualization you can explore the data estimated by UN agencies.
Enter the country – or region – you are interested in in the ‘add country’ box.

# The number of maternal deaths around the world

In 1990 more than half a million women died of causes related to child birth. Since then the number of deaths has declined continuously.

# In the Majority of Countries the Maternal Mortality Rate has recently declined

The visualisation compares the maternal mortality rate in 1990 and one generation later in 2015. The grey line shows where countries would fall in which the rate remained unchanged. Almost all countries lie above the grey line which indicates that the rate was lower in 2015 than in 1990.

# Correlates, Determinants & Consequences

Childbed fever (puerperal sepsis) was a common cause for maternal mortality. It is a bacterial infection of the female reproductive tract  that can occur after giving birth. The introduction of antibiotics in the 1930s made it possible that  in those countries that had access to the new medicine the risk of dying of childbed fever was almost entirely eradicated in the 30s and 40s.

# Maternal mortality rate due to puerperal sepsis, 1861-1900 (Sweden, England & Wales, Scotland) – Gapminder (2010)1 Maternal Mortality Rate due to puerperal sepsis between 1861 and 1900 (Sweden, England Wales, Scotland) – Gapminder (2010)

# Births attended by skilled health staff (% of total) by World Region, average of most recent year available for 2007-2011 – World Development  Indicators (2013)2 Births attended by skilled health staff, average of most recent year available for 2007–11, by World Region – World Development Indicators (2013)

# Maternal mortality reduces education for children

For Kenya Evans and Miguel (2007)3 estimated that school participation falls by 5.5 percentage points immediately following the death of a parent. This decrease appears to be driven by the death of the mother: the post-death decrease in school participation to be 9.3 percentage points and the pre-death decrease to be 6.5 percentage points.

Beegle and Adhvaryu (2012)4 find that in Tanzania, children who lose a mother before turning 15, on average, complete one less year of schooling than other children.

Gourlay et al (2014)5 find that in Zimbabwe, female double-orphans (girls that have lost both parents) are 13 percentage points less likely to be enrolled in school than non-orphans.

# Economic development and maternal mortality

The change of both average income (GDP per capita) and the maternal mortality rate are shown in the following visualisation.
We see a strong correlation, but not every country that achieved economic growth did also achieve a reduction of maternal mortality.

# Data Quality & Definition

The UN Maternal Mortality Estimation Inter-agency Group publishes Trends in maternal mortality: 1990 to 2015. These data are published by the World Bank here.
According to the report the data comes from several sources. In the best cases data from the civil registration systems were used directly to calculate the estimates of maternal mortality rates (MMRs). When these data were not available two-part multi-level regression model were used to estimate MMRs for all target years. The three selected predictor variables in the regression model are: GDP, the general fertility rate (GFR) and the proportion of skilled attendants at birth.

# Data Sources


# Gapminder
  • Data: Maternal mortality ratio (per 100,000 live births)
  • Geographical coverage: Global – by country
  • Time span: Mostly the last decades but historical time series are available for the following countries: Australia (since 1871), Belgium (since 1851), Denmark (1921-1949 & since 1970), Finland (since 1751), Germany (since 1952), Ireland (since 1871), Japan (since 1935), Malaysia (since 1933), The Netherlands (since 1867), New Zealand (since 1972), Sri Lanka (since 1900), Sweden (since 1751), United Kingdom (since 1847), United States (since 1900) – this data is visualized above in this data entry.
  • Available at: Online at
  • The various original sources of this data are documented by Gapminder here.
  • Gapminder also includes data on ‘Births attended by skilled staff’.

# World Development Indicators (WDI) – World Bank
  • Data: Maternal mortality ratio (modeled estimate, per 100,000 live births)’
  • Geographical coverage: Global – by country and world region.
  • Time span: Since 1990 – observations every 5 years for almost all countries.
  • Available at: Online here.
  • Related data in the WDI is available on ‘Pregnant women receiving prenatal care (%)’ and ‘Births attended by skilled health staff (% of total)’.