Can you trust literacy rate?

It does need to be understood with caution. The basic question as defined by the UN and others for statistical purposes is often, "can the person read a typical sentence or two aloud?"

They don't say anything about how easily, or quickly that is done. Or whether it can be done with more than one sample sentence, even. There is of course an incentive to use the loosest definition and not test too aggressively. A 100% literacy rate sounds wonderful, after all. Why, even North Korea has a 100% literacy rate!

Here in Canada where a 99% literacy rate used to be claimed, the loosest sense of that does seem approximately true. Nearly everyone (95%+) can read signs and fill out simple government forms, and perhaps shop online. But it's certainly not 99% that can read quickly and easily. Many people do struggle. They tend to be averse to long stretches like long articles or a whole book. Sometimes it is a learning or intellectual difficulty. Sometimes it is just a lack of practice. I don't know all the factors.

We now use a 5 category literacy test here. I took one version of it in high school. 46% of adults tested like that score in the bottom two categories. Assuming the test is valid, that means they could not understand a typical short story and answer questions about what happened, or read newspaper and magazine articles and write a short essay giving their opinions or response on the topics presented. Most of that group may be literate, but they are not functionally literate.