1. Chapter 2 Economic Methods and Economic Questions

2. Chapter 2 Outline 2.1 The Scientific Method
2 Economic Methods and Economic Questions
Chapter 2 Outline
2.1 The Scientific Method
2.2 Causation and Correlation
2.3 Economic Questions and Answers
Key Ideas
A model is a simplified description of reality.
Economists use data to evaluate the accuracy of models and understand how the world works.
Correlation does not imply causality.
Experiments help economists to measure cause and effect.
Economic research focuses on questions that are important to society and can be answered with models and data.

3. Evidence-Based Economics Example Is college worth it?
2 Economic Methods and Economic Questions
Evidence-Based Economics Example Is college worth it?
The Scientific Method
The scientific method (also referred to as empiricism) is composed of two steps: 1. Developing models that explain some part of the world 2. Testing those models using data to see how closely the model matches what we actually observe

4. Does it look like anyone you know?
2.1 The Scientific Method
Models and Data
Does this it looks like anyone you know? It doesn’t I presume. This as an example of a model. It represents something, but very simply—it abstracts from reality.
What is this?
Does it look like anyone you know?

5. Model - A simplified description of reality Is this an airplane?
2.1 The Scientific Method Models and Data
Model - A simplified description of reality
Is this an airplane?
Wright brothers. Everyone else who was trying to make an airplane that could actually fly built real airplanes, tried to get them up in the air, but they crashed. The reason the Wright brothers were successful is because they built a wind tunnel and built models of airplanes, tested different designs until they found one that worked in the tunnel. Then they built an actual plane. The point is that models are not meant to be perfect replications of the real world. They do give us insights into what is likely to happen. They are meant to convey an idea as simply as possible while still representing the essence of the idea.

6. What’s the shortest distance between two points?
2.1 The Scientific Method
Models and Data
What’s the shortest distance between two points?
So flight paths, when viewed on flat maps, look much more like parabolas than straight lines
Exhibit 2.1 Flying from New York to Tokyo Requires More Than a Flat Map

7. 2.1 The Scientific Method
An Economic Model
Evidenced-Based Example: Returns to education
Assumption—one more year of education results in a 10% increase in future earnings
Returns to education:
If you would earn $15 per hour with 12 years of education, with one more year of education (your first year of college) you would earn:
$15 x 1.10 = $16.50
If you would earn $16.50 with 13 years of education, with one more year of education (2nd year of college), you would earn:
$16.50 x 1.10 = $18.15

8. Returns to education: Hypothesis:
2.1 The Scientific Method
An Economic Model
Returns to education:
The third year: $18.15 x 1.1 = $19.97
The fourth year: $19.97 x 1.1 = $21.97
Returns to education: Hypothesis:
Getting a college degree (years 13-16) increases wages from $15 to $21.97, or 46.5%
[(($ $15)/$15) x 100 = ]
$18.15 x 1.1 = $ ≈ $19.97 and $19.97 x 1.1 = $ ≈ $ You could also follow the text and use the factor of The math might be difficult for some students so the alternative is presented here.

9. Two important features of models:
2.1 The Scientific Method
An Economic Model
Two important features of models:
They are not exact. Not everyone will see his or her wages increase by 10% with every additional year of education
They generate predictions that can be tested with data
Now ready to answer the evidence-based economics question

10. 2.1 The Scientific Method An Economic Model
Hypothesis: Each additional year of education increases wages by 10% True or False?
Exhibit 2.3 Average Annual Earnings of 30-Year-Old Americans by Education Level (2013 data)

11. College results in a wage that is 57% higher.
2.1 The Scientific Method: An Economic Model
How much higher is the wage for college graduates than for high school graduates?
College = $51,780
High School = $32,941
College results in a wage that is 57% higher.
Model predicted 1.46%. Is that close enough?
Models are not going to predict reality exactly. In this case, the model predicted wages that are 46% higher; the data show they are 57% higher. This is a model that is a good approximation of reality in this case, although in other applications, we might prefer a smaller difference (for example, in the case of drug trials).

12. 2.1 The Scientific Method: An Economic Model
If college graduates earn, on average, $51,780/year, does that mean that all college graduates earn that much?
Note: data that are used for many kinds of model-testing are based on averages, or means.
Emphasize that the data that are used for many kinds of model-testing are based on averages, or means. You can show an example computation or go to the next slide to bring the point home that means are often not illustrative of the whole.

13. Can Bill Gates make you rich?
2.1 The Scientific Method
An Economic Model
Can Bill Gates make you rich?
Note: But when it’s skewed (a few at the high or low end), the mean is not a very accurate measure. Thus, the entry of Gates into this class will NOT make you all millionaires.
How does his level of education affect his income?
Note: Bill Gates is a college drop-out. Does that mean that the returns to education model is worthless? No, because that observation—that dropping out of college leads to billionaire status—is based on one person. Conclusions based on small samples are suspect.

14. 2.2 Causation and Correlation
Study from England, showed that cows who were named gave more milk than unnamed cows.
Buttercup
The second study found that gamblers over 65 years of age are healthier than non-gamblers of the same age group.
When these studies were originally released, the news reports strongly implied that the one caused the other: that naming caused greater milk production; that gambling caused increased health among older people. Correlated BUT causal!!

15. 2.2 Causation and Correlation
Causation --When one thing directly affects another
Example: pulling an all-nighter will make you tired
Correlation -- When two things are related
Positive correlation – they both change in the same direction
Negative correlation – they change in opposite directions
Example: shorter skirt lengths are associated with good economic conditions
Correlation - When two things are related
The two previous examples—are they examples of causation or correlation?
Ask students for reasons that naming a cow and milk production might be correlated. Gambling and healthy older people?
For the data on hemlines and economic conditions, go to

16. 2.2 Causation and Correlation
Why isn’t correlation the same thing as causality?
Omitted variables, e.g. buttercup
If we ignore something that contributes to cause and effect, then that something is an omitted variable. A correlation might not make sense until the omitted variable is added.
2. Reverse causality = reverse causality is when there is cause and effect, but it goes in the opposite direction as what we thought.
Example: gambling and healthier older people
The example of gambling and healthier older people is an example of reverse causality. Instead of gambling causing increased health among older folks, the causality runs the other way—increased health among older people causes more of all activities outside the home, including gambling.

17. 2.2 Causation and Correlation
How can we tell the difference between causality and correlation?
Experiments
Controlled = subjects are randomly put into treatment (something happens) and control (nothing happens) groups by the researcher.
Problem: difficult to do with economics studies
Example: naming the cows and looking at milk production is an example of a controlled experiment. The researcher can randomly assign cows into groups that get named, get pampered and see what happens to milk production. This is difficult to do with people, though, since something similar could actually be cruel. In order to understand if the treatment is really effective, however, everything needs to be the same between the two groups except the treatment. Then, any differences between the two groups can be attributed to the treatment.

18. How can we tell the difference between causality and correlation?
2.2 Causation and Correlation
How can we tell the difference between causality and correlation?
Experiments
Natural = subjects end up in treatment or control groups due to something that is not purposefully determined by the researcher
Example: observing whether healthy older people engage in more gambling behavior is a natural experiment. The researcher observes the outcomes, but doesn’t direct the subjects into certain behaviors. Therefore the “observation” is done through analysis of data, not direct observation of the subjects.

19. Evidence-Based Economics and Natural Experiments
2.2 Causation and Correlation
Evidence-Based Economics and Natural Experiments
How much is an extra year of school worth?
In 1947, the U.K. raised the minimum drop-out age from 14 to 15.
Those students reaching age 14 before 1947 = control group
Those students reaching age 14 in 1947 or after = treatment group
Note: researchers didn’t assign students into these two groups—the assignment of subjects into these groups was beyond researchers’ control. On average, staying in school an extra year increased earnings by 10%, which is the basis of the assumption we made in the beginning of the chapter.

20. Two Properties of a Good Economic Question: Relevant and important
2.3 Economic Questions and Answers
Two Properties of a Good Economic Question:
Relevant and important
-- Economic research contributes to social welfare
Social welfare is a common economic term, meaning the greater good, creating an outcome that maximizes the well-being of the greatest number of people.
2. Can be answered
-- Economic questions can be answered empirically