Many people around the world think it is necessary to believe in God to be a moral person, according to surveys in 40 countries by the Pew Research Center. However, this view is more common in poorer countries than in wealthier ones.

Is belief in God essential to morality?

In 22 of the 40 countries surveyed, clear majorities say it is necessary to Pew Research Center to be moral and have good values. This position is highly prevalent, if not universal, in Africa and the Middle East. At least three-quarters in all six countries surveyed in Africa say that faith in God is essential to morality. In the Middle East, roughly seven-in-ten or more agree in Egypt, Jordan, Turkey, the Palestinian territories, Tunisia and Lebanon. Across the two regions, only in Israel does a majority think it is notnecessary to believe in God to be an upright person.

Many people in Asia and Latin America also link faith and morality. For example, Indonesians, Pakistanis, Filipinos and Malaysians almost unanimously think that belief in God is central to having good values. People in El Salvador, Brazil, Bolivia and Venezuela overwhelmingly agree. However, most Chinese take the opposite position – that it is not necessary to be a believer to be a moral person. And in Latin America, the Chileans and Argentines are divided.

In North America and Europe, more people agree that it is possible to be non-religious and still be an upright person. At least half in nearly every country surveyed take this view, including roughly eight-in-ten or more in France, Spain, the Czech Republic and Britain. In these two regions, Americans are unique – 53% say belief in God is necessary to be moral.

These are among the main findings of Pew Research Center surveys conducted among 40,080 people in 40 countries between 2011 and 2013 (see “Survey Methods” for more details). The survey also finds that publics in richer nations tend to place less emphasis on the need to believe in God to have good values than people in poorer countries do. Two countries, however, stand out as clear exceptions to this pattern: the U.S. and China. Americans are much more likely than their economic counterparts to say belief in God is essential to morality, while the Chinese are much less likely to do so.

 

Views on faith and morality vary by age and educationThere are also significant divides within some countries based on age and education, particularly in Europe and North America. In general, individuals age 50 or older and those without a college education are more likely to link morality to religion. For example, in Greece, 62% of older adults say it is necessary to believe in God to be a moral person, while just 29% of 18- to 29-year-olds agree. In the U.S., a majority of individuals without a college degree (59%) say faith is essential to be an upright person, while fewer than four-in-ten college graduates say the same (37%).

Richer Nations Are Exception

Survey Methods

Results for the surveys are based on telephone and face-to-face interviews conducted under the direction of Princeton Survey Research Associates International. Survey results are based on national samples. For further details on sample designs, see below.

The descriptions below show the margin of sampling error based on all interviews conducted in that country. For results based on the full sample in a given country, one can say with 95% confidence that the error attributable to sampling and other random effects is plus or minus the margin of error. In addition to sampling error, one should bear in mind that question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of opinion polls.

Country: Argentina
Sample design: Multi-stage cluster sample stratified by locality size
Mode: Face-to-face adults 18 plus
Languages: Spanish
Fieldwork dates: March 6 – March 26, 2013
Sample size: 819
Margin of Error: +/-4.7 percentage points
Representative: Adult population (excluding dispersed rural population, or 8.8% of the population)
Country: Australia
Sample design: Random Digit Dial (RDD) probability sample of landline and cell phone households
Mode: Telephone adults 18 plus
Languages: English
Fieldwork dates: March 4 – March 18, 2013
Sample size: 800
Margin of Error: +/-4.4 percentage points
Representative: Telephone households (roughly 98% of all Australian households)
Country: Bolivia
Sample design: Multi-stage cluster sample stratified by department and urbanity
Mode: Face-to-face adults 18 plus
Languages: Spanish
Fieldwork dates: March 12 – April 18, 2013
Sample size: 800
Margin of Error: +/-4.5 percentage points
Representative: Adult population (excluding dispersed rural population, or 10% of the population)
Country: Brazil
Sample design: Multi-stage cluster sample stratified by Brazil’s five regions and size of municipality
Mode: Face-to-face adults 18 plus
Languages: Portuguese
Fieldwork dates: March 4 – April 21, 2013
Sample size: 960
Margin of Error: +/-4.1 percentage points
Representative: Adult population
Country: Britain
Sample design: Random Digit Dial (RDD) probability sample representative of all telephone households (roughly 99% of all British households)
Mode: Telephone adults 18 plus
Languages: English
Fieldwork dates: March 22 – April 13, 2011
Sample size: 1,000
Margin of Error: +/-3.5 percentage points
Representative: Telephone households (including cell phone only households)
Country: Canada
Sample design: Random Digit Dial (RDD) probability sample of landline and cell phone-only households
Mode: Telephone adults 18 plus
Languages: English, French
Fieldwork dates: March 5 – March 18, 2013
Sample size: 701
Margin of Error: +/-3.7 percentage points
Representative: Telephone households (excluding residents of Yukon, Nunavut, and Northwest Territories; sample represents roughly 98% of all Canadian households)
Country: Chile
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: Spanish
Fieldwork dates: March 4 – March 19, 2013
Sample size: 800
Margin of Error: +/-5.2 percentage points
Representative: Adult population (excluding Chiloe and other islands, or 3% of the population)
Country: China
Sample design: Multi-stage cluster sample stratified by China’s three regional-economic zones and urbanity. Twelve cities, 12 towns and 12 villages were sampled covering central, east, and west China.
Mode: Face-to-face adults 18 plus
Languages: Chinese (Mandarin, Hebei, Shandong, Yunnan, Chongqing, Guangdong, Hubei, Henan, Hunan, Jiangsu, Gandu, Sichuan, Shaanxi, Anhui, Shanghai, Jilin, Jiangxi, Zhejiang, and Beijing dialects)
Fieldwork dates: March 4 – April 6, 2013
Sample size: 3,226
Margin of Error: +/-3.5 percentage points
Representative: Adult population (excluding Tibet, Xinjiang, Hong Kong and Macau, or roughly 2% of the population). Disproportionately urban. The data were weighted to reflect the actual urbanity distribution in China.
Note: The results cited are from Horizonkey’s self-sponsored survey.
Country: Czech Republic
Sample design: Random Digit Dial (RDD) probability sample of adults who own a cell phone
Mode: Telephone adults 18 plus
Languages: Czech
Fieldwork dates: March 4 – March 14, 2013
Sample size: 700
Margin of Error: +/-3.7 percentage points
Representative: Adults who own a cell phone (roughly 91% of adults age 18 and older)
Country: Egypt
Sample design: Multi-stage cluster sample stratified by governorates and urbanity
Mode: Face-to-face adults 18 plus
Languages: Arabic
Fieldwork dates: March 3 – March 23, 2013
Sample size: 1,000
Margin of Error: +/-4.3 percentage points
Representative: Adult population (excluding Frontier governorates, or about 2% of the population)
Country: El Salvador
Sample design: Multi-stage cluster sample stratified by department and urbanity
Mode: Face-to-face adults 18 plus
Languages: Spanish
Fieldwork dates: April 18 – May 1, 2013
Sample size: 792
Margin of Error: +/-5.3 percentage points
Representative: Adult population
Country: France
Sample design: Random Digit Dial (RDD) sample representative of all telephone households (roughly 99% of all French households) with quotas for gender, age and occupation and proportional to region size and urban/rural population
Mode: Telephone adults 18 plus
Languages: French
Fieldwork dates: March 21 – April 5, 2011
Sample size: 1,004
Margin of Error: +/-3.5 percentage points
Representative: Telephone households (including cell phone only households)
Country: Germany
Sample design: Random Last Two Digit Dial (RL(2)D) probability sample representative of roughly 95% of the German population proportional to population size
Mode: Telephone adults 18 plus
Languages: German
Fieldwork dates: March 21 – April 11, 2011
Sample size: 1,001
Margin of Error: +/-4.5 percentage points
Representative: Telephone households (excluding cell phone only households — between 5% and 10%)
Country: Ghana
Sample design: Multi-stage cluster sample stratified by region and settlement size
Mode: Face-to-face adults 18 plus
Languages: Akan (Twi), English, Dagbani, Ewe
Fieldwork dates: March 20 – April 3, 2013
Sample size: 799
Margin of Error: +/-4.7 percentage points
Representative: Adult population
Country: Greece
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: Greek
Fieldwork dates: March 4 – March 27, 2013
Sample size: 1,000
Margin of Error: +/-3.7 percentage points
Representative: Adult population (excluding the islands in the Aegean and Ionian Seas, or roughly 6% of the population)
Country: Indonesia
Sample design: Multi-stage cluster sample stratified by province and urbanity
Mode: Face-to-face adults 18 plus
Languages: Bahasa Indonesian
Fieldwork dates: March 9 – March 27, 2013
Sample size: 1,000
Margin of Error: +/-4.0 percentage points
Representative: Adult population (excluding Papua and remote areas or provinces with small populations, or 12% of the population)
Country: India
Sample design: Area-probability design. The primary sampling units were urban settlements and rural districts covering 15 of the 17 most populous states (Kerala and Assam were excluded) and the Union Territory of Delhi
Mode: Face-to-face adults 18 plus
Languages: Hindi, Tamil, Bengali, Telugu, Odia, Marathi, Kannada, and Gujarati
Fieldwork dates: December 7, 2013 – January 12, 2014
Sample size: 2,464
Margin of Error: +/-3.8 percentage points
Representative: Proportional allocation of 1,876 interviews by region and urbanity, plus an urban over-sample of 588 interviews. The full sample was weighted to reflect the national urban-rural distribution in India. Sample covers roughly 91% of the Indian population
Country: Israel
Sample design: Multi-stage cluster sample stratified by Israel’s six districts, urbanity, and socioeconomic status, with an oversample of Arabs
Mode: Face-to-face adults 18 plus
Languages: Hebrew, Arabic
Fieldwork dates: March 29 – April 12, 2013
Sample size: 922 (504 Jews, 406 Arabs, 12 others)
Margin of Error: +/-4.6 percentage points
Representative: Adult population
Country: Italy
Sample design: Multi-stage cluster sample stratified by four regions and urbanity
Mode: Face-to-face adults 18 plus
Languages: Italian
Fieldwork dates: March 4 – March 19, 2013
Sample size: 1,105
Margin of Error: +/-4.1 percentage points
Representative: Adult population
Country: Japan
Sample design: Random Digit Dial (RDD) probability sample representative of all landline telephone households stratified by region and population size (excluding 5.4% of the population living in areas most affected by the March 11, 2011 earthquake and tsunami)
Mode: Telephone adults 18 plus
Languages: Japanese
Fieldwork dates: April 8 – April 27, May 13 – May 24, 2011
Sample size: 700
Margin of Error: +/-4.5 percentage points
Representative: Telephone households (excluding cell phone only households – less than 5%, households with no telephones – about 5%, and the population living in areas most affected by the March 11, 2011 earthquake and tsunami – 5.4%)
Country: Jordan
Sample design: Multi-stage cluster sample stratified by Jordan’s 12 governorates and urbanity
Mode: Face-to-face adults 18 plus
Languages: Arabic
Fieldwork dates: March 4 – March 23, 2013
Sample size: 1,000
Margin of Error: +/-4.5 percentage points
Representative: Adult population
Country: Kenya
Sample design: Multi-stage cluster sample stratified by province and settlement size
Mode: Face-to-face adults 18 plus
Languages: Kiswahili, English
Fieldwork dates: March 13 – March 30, 2013
Sample size: 798
Margin of Error: +/-4.3 percentage points
Representative: Adult population
Country: Lebanon
Sample design: Multi-stage cluster sample stratified by Lebanon’s seven regions and urbanity
Mode: Face-to-face adults 18 plus
Languages: Arabic
Fieldwork dates: March 4 – March 22, 2013
Sample size: 1,000
Margin of Error: +/-4.0 percentage points
Representative: Adult population (excluding a small area in Beirut controlled by a militia group and a few villages in the south of Lebanon, which border Israel and are inaccessible to outsiders, or about 2% of the population)
Country: Malaysia
Sample design: Multi-stage cluster sample stratified by state and urbanity
Mode: Face-to-face adults 18 plus
Languages: Malay, Mandarin Chinese, English
Fieldwork dates: March 4 – April 3, 2013
Sample size: 822
Margin of Error: +/-4.3 percentage points
Representative: Adult population (excluding difficult to access areas in Sabah and Sarawak, or about 7% of the population)
Country: Mexico
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: Spanish
Fieldwork dates: March 4 – March 17, 2013
Sample size: 1,000
Margin of Error: +/-4.1 percentage points
Representative: Adult population
Country: Nigeria
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: English, Hausa, Yoruba, Igbo
Fieldwork dates: March 6 – April 4, 2013
Sample size: 1,031
Margin of Error: +/-4.0 percentage points
Representative: Adult population (excluding Borno, Yobe and some areas in Taraba, or about 5% of the population)
Country: Pakistan
Sample design: Multi-stage cluster sample stratified by province and urbanity
Mode: Face-to-face adults 18 plus
Languages: Urdu, Pashto, Punjabi, Saraiki, Sindhi
Fieldwork dates: March 11 – March 31, 2013
Sample size: 1,201
Margin of Error: +/-4.3 percentage points
Representative: Adult population (excluding the Federally Administered Tribal Areas, Gilgit-Baltistan, Azad Jammu and Kashmir for security reasons as well as areas of instability in Khyber Pakhtunkhwa [formerly the North-West Frontier Province] and Baluchistan, or roughly 18% of the population). Disproportionately urban. The data were weighted to reflect the actual urbanity distribution in Pakistan.
Country: Palestinian territories
Sample design: Multi-stage cluster sample stratified by region and urban/rural/refugee camp population
Mode: Face-to-face adults 18 plus
Languages: Arabic
Fieldwork dates: March 29 – April 7, 2013
Sample size: 810
Margin of Error: +/-4.4 percentage points
Representative: Adult population (excluding Bedouins who regularly change residence and some communities near Israeli settlements where military restrictions make access difficult, or roughly 5% of the population)
Country: Philippines
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: Tagalog, Cebuano, Ilonggo, Ilocano, Bicolano
Fieldwork dates: March 10 – April 3, 2013
Sample size: 804
Margin of Error: +/-4.5 percentage points
Representative: Adult population
Country: Poland
Sample design: Multi-stage cluster sample stratified by Poland’s 16 provinces and urbanity
Mode: Face-to-face adults 18 plus
Languages: Polish
Fieldwork dates: March 2 – March 24, 2013
Sample size: 800
Margin of Error: +/-3.9 percentage points
Representative: Adult population
Country: Russia
Sample design: Multi-stage cluster sample stratified by Russia’s eight regions plus Moscow and St. Petersburg and urbanity
Mode: Face-to-face adults 18 plus
Languages: Russian
Fieldwork dates: March 5 – March 21, 2013
Sample size: 996
Margin of Error: +/-3.6 percentage points
Representative: Adult population (excluding High North regions, the Chechen Republic, and the Ingush Republic, or about 3% of the population)
Country: Senegal
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: Wolof, French
Fieldwork dates: March 6 – March 30, 2013
Sample size: 800
Margin of Error: +/-4.1 percentage points
Representative: Adult population
Country: South Africa
Sample design: Multi-stage cluster sample stratified by metropolitan area, province and urbanity
Mode: Face-to-face adults 18 plus
Languages: English, Zulu, Xhosa, South Sotho, Afrikaans
Fieldwork dates: March 18 – April 12, 2013
Sample size: 815
Margin of Error: +/-4.1 percentage points
Representative: Adult population
Country: South Korea
Sample design: Random Digit Dial (RDD) probability sample of adults who own a cell phone
Mode: Telephone adults 18 plus
Languages: Korean
Fieldwork dates: March 4 – March 18, 2013
Sample size: 809
Margin of Error: +/-3.7 percentage points
Representative: Adults who own a cell phone (roughly 96% of adults age 18 and older)
Country: Spain
Sample design: Random Digit Dial (RDD) probability sample representative of telephone households (about 99% of Spanish households) stratified by region and proportional to population size
Mode: Telephone adults 18 plus
Languages: Spanish/Castilian
Fieldwork dates: March 22 – April 5, 2011
Sample size: 1,000
Margin of Error: +/-3.5 percentage points
Representative: Telephone households (including cell phone only households)
Country: Tunisia
Sample design: Multi-stage cluster sample stratified by governorate and urbanity
Mode: Face-to-face adults 18 plus
Languages: Tunisian Arabic
Fieldwork dates: March 4 – March 19, 2013
Sample size: 1,000
Margin of Error: +/-4.0 percentage points
Representative: Adult population
Country: Turkey
Sample design: Multi-stage cluster sample stratified by the 26 regions (based on geographical location and level of development (NUTS 2)) and urbanity
Mode: Face-to-face adults 18 plus
Languages: Turkish
Fieldwork dates: March 5 – March 24, 2013
Sample size: 1,000
Margin of Error: +/-7.7 percentage points
Representative: Adult population
Country: Uganda
Sample design: Multi-stage cluster sample stratified by region and urbanity
Mode: Face-to-face adults 18 plus
Languages: Luganda, English, Runyankole/Rukiga, Luo, Runyoro/Rutoro, Ateso, Lugbara
Fieldwork dates: March 15 – March 29, 2013
Sample size: 800
Margin of Error: +/-4.3 percentage points
Representative: Adult population
Country: United States
Sample design: Random Digit Dial (RDD) probability sample representative of all telephone households in the continental U.S. stratified by county
Mode: Telephone adults 18 plus
Languages: English
Fieldwork dates: March 25 – April 14, 2011
Sample size: 1,001
Margin of Error: +/-4.0 percentage points
Representative: Telephone households in continental U.S. (including cell phone only households)
Country: Venezuela
Sample design: Multi-stage cluster sample stratified by region and parish size
Mode: Face-to-face adults 18 plus
Languages: Spanish
Fieldwork dates: March 15 – April 27, 2013
Sample size: 1,000
Margin of Error: +/-3.5 percentage points
Representative: Adult population (excluding remote areas, or about 4% of population)

(Pew Research Center)