Roland: The alarming fact is that 90 percent of the people in this cou

The heart of Sharon's argument is pretty straightforward: she thinks that what Roland sees is somewhat "normal", and not "alarming".

Let’s break down the reasoning behind her argument:

1. A "normal" unemployment rate is 1/20.
2. So if you know 20 typical workers, odds are good that one will be unemployed.
3. And then if you know 50 workers, at least 1 of them will probably be unemployed.
4. Therefore, it’s likely at any given time that 90% of people in the country know at least 1 unemployed person.

In her reasoning, Sharon refers to nationwide levels of unemployment. When she jumps to step 4 of her argument, she assumes that the employment patterns of the 50 workers each of us knows personally will resemble the nationwide employment patterns. In order to accept this assumption, we need evidence that the normal unemployment rate in any given area will roughly match the normal rate of unemployment for the entire country. Otherwise, it could be the case that the unemployed workers are overwhelmingly concentrated in a few parts of the country, and most people elsewhere might NOT know any unemployed workers.

So Sharon's argument relies on which of these assumptions?


It makes no difference whether normal levels of unemployment are exceeded rarely or frequently. As long as the current level of unemployment is normal, then Sharon’s argument is valid.

In other words, normal levels of unemployment could be exceeded frequently. But according to Sharon, the data cited by Roland is evidence that unemployment levels are normal right now. Sharon’s argument does not rely on choice (A), so eliminate this one.


Choice (B) gets to the heart of Sharon’s assumption. If unemployment is evenly distributed across the population as opposed to being concentrated in certain states, cities and industries, then we’ll have an easier time agreeing with Sharon. If (B) is true, then any person who knows approximately 50 workers -- anywhere in the country -- is likely to know at least one unemployed worker, even if unemployment levels are moderate.

If (B) were NOT true and unemployment levels were moderate, then we would expect people in the geographically isolated segments to know several unemployed workers. In that case, most people in other parts of the country would NOT likely know at least one unemployed worker. If (B) were not true, then Roland’s evidence would be "alarming", and Sharon’s argument would fall apart.

Let’s keep choice (B) for now and try to eliminate the rest.


Choice (C) could certainly weaken Roland’s argument (by suggesting that the evidence is normal, not alarming). But does Sharon’s argument rely on this assumption? What if unemployment levels are sometimes LESS than moderate? In that case, there would certainly be times when the number of people who each know someone who is unemployed is NOT higher than 90% of the population.

Regardless, we are only interested in the fact that that 90 percent NOW report that they know someone who is unemployed. According to Sharon, this is no cause for alarm. Sharon’s argument would be the same regardless of whether (C) is true, so we can eliminate this one.


Sharon’s argument doesn’t depend on whether Roland is honest. There would be no reason for continuing to this conversation if Roland were rattling off fake news he saw on Facebook, but Sharon’s logical connection wouldn’t be affected.

Sharon’s argument is basically, “Even if your evidence is true, there is no cause for alarm.” If the evidence is false, Roland might be a liar, but Sharon’s logic remains sound.

(D) isn't necessary, so we can eliminate it.


Choice (E) very well may be true, but it tells us nothing that would affect the logical argument Sharon is making. She says we shouldn’t be alarmed because normal unemployment rates explain the seemingly abnormal rates of knowing an unemployed person. Fear of losing one's job is completely irrelevant to her argument.

So (E) is out, and (B) is our answer.