Fact Check

"The UN had predicted global mortality rate for the years 2015 to 2020 of 8.1"

The UN World Population Prospects data does not support a figure of 8.1 per 1,000 for the 2015–2020 period. The 2024 revision estimates the world crude death rate for 2014–2016 at approximately 7.5 per 1,000. The figure of 8.1 may have been confused with the 2011 actual rate (~8.12 from IndexMundi/CIA data), or possibly an older revision's projection. The number appears conflated from a different source or time period.

"Based on CIA World Factbook data for 2016, the figure is estimated at 7.78. This was down from 7.89 in 2014."

The 2014 figure of 7.89 is confirmed — IndexMundi lists the world death rate for 2014 as exactly 7.89 deaths per 1,000 population. However, the 2016 figure is slightly off: the published CIA/IndexMundi figure is 7.7, not 7.78. The overall trend (declining from 7.89 to approximately 7.7) is correct, but the specific number cited is wrong.

The article argues 2016 was not exceptional for celebrity deaths — just confirmation bias amplified by online content creators.

A defensible argument, but contested by data. Jason Crease's empirical analysis (hosted by UC Berkeley's statistics department) found that among the top 200 most famous people by Wikipedia page length, 25 died in 2016 versus a predicted 17 — roughly a once-in-200-years event. The BBC's obituaries editor confirmed a 50% increase in pre-prepared obituaries used in 2016 vs 2015. However, Brogan's argument about confirmation bias and media amplification is also well-supported. The truth is likely both: 2016 was genuinely anomalous AND the perception was amplified by social media.

"This basically reflects the percentage of the number of days in each month."

This is a well-known methodological point in epidemiology and demography, formally called the "length-of-month effect." Months with 31 days naturally record more events than months with 28–30 days. Researchers routinely adjust monthly counts by multiplying by (30.44 / actual days in month) to standardise for comparison. Brogan is correct to flag it as a confounding variable that laypeople overlook.

"December and January which have amplified positions on the curve, compared against the month sizes."

Winter excess mortality is one of the most robustly documented phenomena in epidemiology. Death rates in the Northern Hemisphere peak in December, January, and February, driven primarily by cardiovascular disease (peaks late February) and respiratory disease (peaks late January). Studies show winter mortality can be 20–35% higher than summer baseline. This effect persists after day-count adjustment — it is genuine seasonality, not an artefact of month length.

"The number of notable deaths has been rising since 2006."

Directionally supported but confounded by at least three factors: population growth means more people dying; the generation that became famous via mass media (1950s–1970s) is reaching natural mortality age; and Wikipedia itself has grown enormously since 2006, meaning more deaths are documented now than before. The underlying demographic argument is sound and widely cited, but attributing the trend purely to "more notable deaths occurring" without acknowledging Wikipedia's own growth overstates the case.