Volume 53 - Article 22 | Pages 661–704  

Online obituaries as a complementary source of data for mortality in Canada

By Pietro Violo, Nadine Ouellette

Abstract

Background: Obituaries and death notices have existed for centuries as a form of commemoration, particularly in Western countries. With the rise of the internet, these records have become more accessible, presenting a valuable, largely untapped source for mortality research.

Objective: We aim to collect online obituaries through web scraping and evaluate their representa-tiveness, advantages, and limitations for use in mortality studies in Canada’s two largest provinces: Quebec and Ontario.

Methods: We web scraped 236,290 and 288,623 obituaries for Quebec and Ontario, respectively, spanning the years 2017 to 2022. Using regular expressions, a formal language for defining text-search patterns, we derived demographic variables from the text to compute mortality measures, which we then compared to a gold-standard vital statistics dataset.

Results: Although obituaries in Quebec and Ontario respectively account for only half and one-third of all recorded deaths, the age and gender distributions they capture closely align with those of the general population. Infant deaths remain notably underrepresented. Life expectancy estimates derived from obituaries exceed official figures by 0.4 years for women and 0.5 for men, while the modal age at death is slightly underestimated. Despite these limitations, the timeliness and demographic representativeness of online obituaries make them a valuable supplement to conventional mortality datasets in Canada.

Contribution: This study draws attention to an underused data source by leveraging Canada’s bilingual context and developing methods for extracting demographic information from both French and English obituaries. We contribute to digital and computational demography by detailing techniques for web scraping, data cleaning, extraction, and validation, and by assessing coverage, age structures, gender disparities, and inherent biases in this type of textual data.

Author’s Affiliation

• Pietro Violo - Université de Montréal, Canada EMAIL
• Nadine Ouellette - Université de Montréal, Canada EMAIL

Other articles by the same author/authors in Demographic Research

The question of the human mortality plateau: Contrasting insights by longevity pioneers
Volume 48 - Article 11

Insight on 'typical' longevity: An analysis of the modal lifespan by leading causes of death in Canada
Volume 35 - Article 17

Changes in the age-at-death distribution in four low mortality countries: A nonparametric approach
Volume 25 - Article 19

Similar articles in Demographic Research

Bayesian multidimensional mortality reconstruction
Volume 54 - Article 28    | Keywords: Bayesian reconstruction, data lack, hierarchical modelling, mortality

Winter life expectancy reduction in Europe
Volume 54 - Article 26    | Keywords: Europe, excess winter deaths, excess winter mortality paradox, life expectancy, mortality, summer, weekly mortality data, winter

Bringing cause-of-death analysis into demography: An interview with France Meslé
Volume 54 - Article 24    | Keywords: causes of death, epidemiological transition, health transition, mortality, mortality data

Refining seasonal mortality estimates through age adjustment: Evidence from Serbia, 2015–2023
Volume 54 - Article 15    | Keywords: age adjustment, excess mortality, life expectancy, mortality, mortality estimates, seasonal fluctuations, Serbia

Feminicide as a determinant of Mexican female life expectancy in the 21st century
Volume 53 - Article 24    | Keywords: female life expectancy, feminicide, life expectancy, Mexico, mortality, violence, women