IQ tests are not a true representation of society and how smart an individual is and their ability to do certain tasks.
If you use as your yard-stick the Stanford-Binet IQ test, it has long been known for it's skewed results that reflect a racial non-diversity....when comparing minorities and blacks it no longer can be relied upon as accurate.
Also as an example, the level of literacy of North American populace differs greatly from the time period of 1916 to 2016. Just because a tested individual is not as literate, it doesn't necessarily mean they can't be smart. They may be financially disadvantaged and had to leave school early.
Take for example...if you were testing in the year 1916...and did testing for the norm for that era, you might create the test around criteria and tasks such as building a cart, or (due to more farm-oriented culture) create a test that was oriented towards usage of farm implements.
However, if you created the test around vocabulary, advanced math, and academics and sciences, then you will see a skewed and inaccurate result.