I do not think 3 writters is enough to draw any conclusion about the security of Bangladesh. They got 156 million people and 20 writters have died this year out of the 7.125 billion people. That means that 0.5 writters should have died in expecation in Bangladesh. 6 times the expected number does not seem unreasonable.
To be more precise: If we assume that 0.5 is the real expectation for Bangladesh and each writer dies with equal probability in a given year and are independent (i.e. that one writer dies does not say anything about the probability that another dies), we can given an upper bound on the probability that 3 have died in a year using Chernoff bounds. It says that the probability that 3 writers have died while the mean is only 0.5 is a bit more than 5%.
Note though that this is an upper bound on the real probability and hence only gives a quite rough estimate. If we assume that it is accurate, we see that if all countries were the sized of Bangladesh (i.e there are 45 countries), then every year more than two countries should have 3 dead writers in expectation (or Bangladesh should have it once every 20 years if it happens only according to chance).
Note that this is not taking into account the relatively arbitrary choice of time for this article (it is more likely that at some point in a year a country is far from the expecation, simply because there are so many points in time in a year), choice of objective (there are many equally important properties to dead writers this year and it is therefore more likely that some of them are far from the expecation) and that it is in the news (more unusual things are in the news, i.e. it is not news that another country did not have 6 times to many dead writers).