In Chapter 3 we learned how to calculate the crude death rate. We divide the num
In Chapter 3 we learned how to calculate the crude death rate. We divide the number of deaths by the midyear population, then multiply by 1000 for convenience. Because the crude death rate has population in its denominator, it automatically adjusts for the fact that larger populations will have more people dying. There is no reason that a large country must have a higher death rate than a smaller country.
Population size does not directly affect the death rate, but having a lot of older people in your population does.
Japan has an older population than Afghanistan, so old that Japan’s 2019 crude death rate, 11 per 1000, was higher than Afghanistan’s crude death rate, 6 per 1000. This makes Japan look like a less healthy and more dangerous place to live than Afghanistan. Is it?
No. At every age of life, the 2019 age-specific mortality rates were lower in Japan than they were in Afghanistan. A twenty-year old in Afghanistan had a higher chance of dying that year than a twenty-year old in Japan. However, a twenty-year old in Afghanistan did not have a higher chance of dying that year than a fifty-year old in Japan. And Japan had a higher fraction of its population age 50 or older than Afghanistan did. That’s why Japan’s crude death rate was higher than Afghanistan’s, and why the crude death rate can be misleading when comparing countries.
Let’s learn how to transform a crude death rate into a standardized death rate. We’ll use the Table below.
Table 5-1 shows us the actual number of deaths from COVID-19 in Taiwan and Japan as of September 30, 2020. By that date, Taiwan had experienced a total of 7 deaths, 2 from 25-49 year-olds, 4 from the 50-74 age category, and 1 person age 75+. If you divide those 7 deaths by Taiwan’s population of 23,562,323 people, then multiply by 1000, you get a crude death rate for Taiwan of 0.0003 deaths per 1000 people, or 3 deaths for every ten million people. Clearly, Taiwan was doing very well.
Japan wasn’t doing too badly either, with a crude death rate of 125 per ten million people, but its crude death rate was about 40 times as high as Taiwan’s.
Now let’s standardize the death rates. To do so, we are going pretend that each country has the same, standard age distribution. What is this standard age distribution? Often, the United States population is used as the standard. Or we could use the World Health Organization’s “World Standard Population”, which is an estimate of the age distribution of the entire world.
In this example, we are going to choose Taiwan’s age distribution as the standard. Taiwan will be our “reference country”.
Since Taiwan already has Taiwan’s age distribution, we don’t have to do anything to its data. Taiwan’s death rate, standardized to itself, will simply be its original crude death rate of 3 per million.
However, Japan’s standardized death rate will be different from its crude death rate.
Step 1 of finding Japan’s standardized death rate is to find the number of people of each age who would