The importance of numbers have been generally neglected as part of the social sciences. Political Science and other social sciences, also known as arts and sciences or liberal arts, are often viewed as unnecessary to society because of their inapplicability, and obsolete at the expense of business education. People tend to see no job opportunities in these fields, except for Psychology and the like, especially after they witness read disputes on such topics and be able to understand them better than they can understand biology, engineering or other fields where more memorization and drill is required.
An exception to this social trend is Economics where debates are still going on as to whether or not it is a business or a social science major. And for good reason! Economics is oftentimes part of the Schools of Business at the undergraduate level and part of the Schools of Arts and Sciences at the graduate level. M.A. in Economics means Master of Arts in Economics.
No, we can’t blame them
We can’t blame people for having established such attitudes toward social sciences, especially when it comes to disputes everyone has formed an opinion of, such as gay marriage, abortion rights, death penalty, gun control and the like – and few would accept other arguments. Listening to what FOX News or MSNBC have to say about these issues opens one’s eyes about one side of the debate only – whatever the respective entity advocates for – and turns Political Science, for example, into what I would call in this aricle “buy-it-or-not” science: one where there is nothing scientific but the word “science” or the ending “-logy” (which also means science).
What’s missing to the concoction is numbers. In order to prove something, premises and conclusions are often not enough, for even correct premises may bring us to a false conclusion, especially if we don’t consider or don’t pay enough attention to alternatives. Anthony Weston, in his “Rulebook for Arguments” has conceptually one of the best examples as a proof. In Appendix I of his book, where he discusses some common fallacies, he introduced a potentially false conclusion deducted from two true premises:
When the roads are icy, the mail is late.
The roads are not icy.
Therefore, the mail is not late.
Such a course of thought is highly convincing but there may be other reasons for the mail to be late. A debate is born which will hardly ever end unless numbers are provided. Measuring the potential of mail deliveries in times of both icy and non-icy roads takes the issue to the next level but is not enough in jumping to a conclusion. Measuring potential health issues on mail deliverers during cold weather (for roads can’t be icy in a weather with persistent temperatures above zero) and their experience behind the wheel when the roads are icy would reveal to us whether there are other reasons besides the ice on the roads that stalls mail delivery. That’s what social scientists do.
In Political Science…
In Political Science, numbers are important too. An example is electoral politics where no testing of theories would lead to opposite answers. What the constituents think about the issues relevant to the political office being run for by a main candidate, and how similar a main candidate’s positions on these issues is compared to theirs is often of decisive importance in who would win the election.
Here’s an example of two states’ two races for the 2010 midterm elections in the U.S. presented by two polling institutes – Quinnipiac University Poll and Rasmussen Reports. These are the gubernatorial and the senatorial races in Connecticut and Florida:
In the race for Governor in Connecticut, both organizations concluded that the race was too close to call or a toss-up respectively (Quinnipiac Poll and Rasmussen Reports use different terms for one and the same conclusion) – with Republican Tom Foley leading Democrat Dan Malloy 48-45 percent (Q-Poll) with a margin of error of 3.2 percentage points and 49-46 percent (Rasmussen) with a margin of error of 4 percentage points. Neither of the two results shows a total of 100 percentage points. Most of the rest, as shown on the links, stand for don’t know or no answer (DK/NA or Not Sure). In the Connecticut race, the likely voters who responded that they don’t know, or are not sure, or gave no answer are 6 percent of the likely voters in the Quinnipiac survey and 5 percent of the likely voters. These are the group of uninformed voters who would vote in the election. They are not informed due to having not followed the news and/or having not researched enough in order for them to know who to vote for. We have no information of which of the two possible factors is greater. Since they are likely voters, they will vote for either Dan Malloy, or Tom Foley, so arguably the best way to estimate that is to consider all possible combinations of this type of electorate’s voting. Possible combinations range from having all 6 percent or all 5 percent of the respective polls vote for either Dan Malloy or Tom Foley. Thus the Quinnipiac Poll indicates that Dan Malloy might win by 3 percentage points (51-48) or that Tom Foley might win by 9 percentage points (45-54), while the Rasmussen Reports indicates that Dan Malloy might win by three percentage points (51-48) or that Tom Foley might win by 7 percentage points (53-46). The question of whether voters of either of the candidates could change their mind, asked by the Quinnipiac Poll, makes matters even more complicated. Roughly 10 percent of either of the two candidates’ likely voters, who participated in the survey, answered that they could change their mind. The way to read these numbers is 10 percent of Mr. Malloy’s 45 percentage points (4.5 percent) will either end up not changing their mind, or vote for Mr. Foley, or vote for someone else or won’t vote at all. Since the two extremities in the results show a different winner, the race is “too close to call” or a “toss-up.” The concept of the margin of error is how likely it is to get any results outside the aforementioned ranges. Final result of the election: Dan Malloy barely won so the two polls’ numbers turned out to be correct.
In the race for Senate in Connecticut, Richard Blumenthal was the frontrunner with a 53-44 percentage points lead (Quinnipiac Poll) and a 53-46 percentage points lead over Linda McMahon (Rasmussen). We can see that the likely voters who answered unsure are 1 percent in the Rasmussen results which gives Dick Blumenthal a 54-46 percentage points victory or a 53-47 percentage points victory. In the Quinnipiac survey, we can see that the likely voters who answered DK/NA are 3 percent of the likely voters meaning that, following the logic from above, the Q-Poll’s result might range from 56-44 percentage points if all of the unsure voters prefer Richard Blumenthal, and 53-47 percentage points if all of the unsure voters prefer Linda McMahon. Adding the results from the question that asks whether the likely voters – whether Mr. Blumenthal’s supporters or Mrs. MacMahon’s supporters – have made up their mind, 4 percent of Dick Blumenthal’s supporters answered they might change their mind. That would potentially give Linda McMahon 2.6 percentage points in the Q-Poll’s data or, in other words, up to 49.6 percentage points – slightly down from Mr. Blumenthal’s minimum of 50.4 percentage points. Therefore, all aforementioned ranges in the race for the U.S. Senate, show Richard Blumenthal as the winner. Final result of the election: Richard Blumenthal won.
In the race for Governor in Florida, Democrat Alex Sink and Republican Rick Scott were very close to each other, with Alex Sink leading by merely 1 percentage point (44-43) in the Quinnipiac poll, and 48-45 in the Rasmussen poll. The 9 percentage points undecided likely voters, according to the Quinnipiac Poll, and the percentage points of likely voters of either of the two candidates, who said they might change their minds (between 2.6 and 2.7 percentage points), would very likely determine the winner of the race. Final results: Rick Scott barely beat Alex Sin
In the race for Senate in Florida, Republican Marco Rubio had a tremendous lead over, Independent Charlie Christ and Democrat Kendrick Meek – enough so that no undecided or unsure voters would be able to determine who the winner is. Final results: Marco Rubio won the election effortlessly.
Being able to interpret numbers in whatever field you might end up working at is always to your favor. Unlike in philosophical disputes where both sides of the debates could be understood differently based on the way one was raised or based on how one educated themselves on the issue, in empirical evidence (the use of numbers) ideology doesn’t matter. Two plus two always equals four, for example.
So don’t be afraid to calculate, don’t be afraid to become more interested in statistics, don’t be afraid to minor in Economics if you are majoring in social sciences. One time, I asked a fellow Political Scientist why she didn’t minor in Economics, and her answer was that she couldn’t stand Maths. If only she could realize what she is missing! The ability to use data at a higher level in your everyday work as a Political Scientist would make you a better, and therefore more demanded in the job market, Political Scientist.
One of my Economics professors at my university – Donn Johnson – emphasized the importance of quantitative skills in the job market. I couldn’t agree more.