Lies, Damn Lies, and Statistics

1
Lies, Damn Lies, and Statistics

• Using Economic Data

2
Empirical Questions
3
Empirical Questions

• What exactly are you trying to measure? Is your
  variable consistent with what youre trying to
  measure?

4
ExamplePoverty in the US

5
Defining Poverty

• Poverty in the US
• Poverty was defined by Mollie Orshansky of the
  SSA in 1964 as 3 times the cost of the Dept. of
  Agricultures economy food plan

6
Defining Poverty

• Poverty in the US
• Poverty was defined by Mollie Orshansky of the
  SSA in 1964 as 3 times the cost of the Dept. of
  Agricultures economy food plan
• Since 1964, that number has been updated annually
  for changes in inflation

7
Defining Poverty

• Poverty in the US
• Poverty was defined by Mollie Orshansky of the
  SSA in 1964 as 3 times the cost of the Dept. of
  Agricultures economy food plan
• Since 1964, that number has been updated annually
  for changes in inflation
• Currently, the poverty line is 9,359/yr for a
  single person

8
Defining Poverty

• Poverty in the US
• Poverty was defined by Mollie Orshansky of the
  SSA in 1964 as 3 times the cost of the Dept. of
  Agricultures economy food plan
• Since 1964, that number has been updated annually
  for changes in inflation
• Currently, the poverty line is 9,359/yr for a
  single person

• International Poverty
• Of the 184 member countries of the world bank. 52
  countries are considered high income defined
  as a per capita income of more than 9,206/yr

9
Defining Poverty

• Poverty in the US
• Poverty was defined by Mollie Orshansky of the
  SSA in 1964 as 3 times the cost of the Dept. of
  Agricultures economy food plan
• Since 1964, that number has been updated annually
  for changes in inflation
• Currently, the poverty line is 9,359/yr for a
  single person

• International Poverty
• Of the 184 member countries of the world bank. 52
  countries are considered high income per
  capita income of more than 9,206/yr
• 66 countries are considered low income (less
  than 746/yr)

10
Defining Poverty

• Poverty in the US
• Poverty was defined by Mollie Orshansky of the
  SSA in 1964 as 3 times the cost of the Dept. of
  Agricultures economy food plan
• Since 1964, that number has been updated annually
  for changes in inflation
• Currently, the poverty line is 9,359/yr for a
  single person

• International Poverty
• Of the 184 member countries of the world bank. 52
  countries are considered high income per
  capita income of more than 9,206/yr
• 66 countries are considered low income (less
  than 746/yr)
• Currently the international poverty standard is
  1/day

11
Empirical Questions

• What exactly are you trying to measure? Is your
  variable consistent with what youre trying to
  measure?
• How is your variable measured?

12
Example US Unemployment
13
Measuring Unemployment

• Each month, the Department of Labor surveys
  60,000 households. Each household is placed in
  one of four categories

14
Measuring Unemployment

• Each month, the Department of Labor surveys
  60,000 households. Each household is placed in
  one of four categories
• Under 16 or institutionalized

15
Measuring Unemployment

• Each month, the Department of Labor surveys
  60,000 households. Each household is placed in
  one of four categories
• Under 16 or institutionalized
• Choose not to work Not in Labor Force

16
Measuring Unemployment

• Each month, the Department of Labor surveys
  60,000 households. Each household is placed in
  one of four categories
• Under 16 or institutionalized
• Choose not to work Not in Labor Force
• Choose to work and are working Employed

17
Measuring Unemployment

• Each month, the Department of Labor surveys
  60,000 households. Each household is placed in
  one of four categories
• Under 16 or institutionalized
• Choose not to work Not in Labor Force
• Choose to work and are working Employed
• Choose to work, but cant find a job Unemployed

18
Measuring Unemployment

• Each month, the Department of Labor surveys
  60,000 households. Each household is placed in
  one of four categories
• Under 16 or institutionalized
• Choose not to work Not in Labor Force
• Choose to work and are working Employed
• Choose to work, but cant find a job Unemployed
• Unemployment Rate D/(CD)

19
Is the unemployment rate biased downward?
20
Is the unemployment rate biased downward?

• The unemployment rate doesnt count
  underemployment (those that would like to work
  full time, but only work part time)

21
Is the unemployment rate biased downward?

• The unemployment rate doesnt count
  underemployment (those that would like to work
  full time, but only work part time)
• The discouraged worker effect Those that have
  given up trying to find a job are counted as not
  in the labor force rather than unemployed

22
Is the unemployment rate biased upward?
23
Is the unemployment rate biased upward?

• Selection bias those that are unemployed are
  more likely to be home to answer the survey.

24
Is the unemployment rate biased upward?

• Selection bias those that are unemployed are
  more likely to be home to answer the survey.
• Moral hazard due to unemployment insurance, it
  is difficult to tell how hard individuals are
  trying to find work

25
Other Problems

• Should we interpret unemployment statistics
  differently when population demographics change?
  (e.g. individuals under the age of 25 are much
  more likely to be unemployed)

26
Other Problems

• Should we interpret unemployment statistics
  differently when population demographics change?
  (e.g. individuals under the age of 25 are much
  more likely to be unemployed)
• Should we count military personnel as employed or
  unemployed

27
Empirical Questions

• What exactly are you trying to measure? Is your
  variable consistent with what youre trying to
  measure?
• How is your variable measured?
• Is your variable in real or nominal terms?

28
Example Suppose that you have 100 to invest in
either the US or Argentina. Given the current
interest rates, where should you invest?

• Argentina
• i 12.8

• United States
• i 4.25

29
Example Suppose that you have 100 to invest in
either the US or Argentina. Given the current
interest rates, where should you invest?

• Argentina
• i 12.8
• Annual inflation rate 14.3

• United States
• i 4.25
• Annual inflation rate 2.4

30
Example Suppose that you have 100 to invest in
either the US or Argentina. Given the current
interest rates, where should you invest?

• Argentina
• i 12.8
• Annual inflation 14.3
• Real (inflation adjusted) return -1.5

• United States
• i 4.25
• Annual inflation 2.4
• Real (inflation adjusted) return 1.85

31
Real vs. Nominal Variables
32
Real vs. Nominal Variables

• Nominal variables are measured in terms of some
  currency (e.g. your annual income is 70,000 per
  year)

33
Real vs. Nominal Variables

• Nominal variables are measured in terms of some
  currency (e.g. your nominal income is 70,000 per
  year)
• Real (inflation adjusted) variables are measured
  in terms of some commodity (e.g. your real income
  is 7,000 pizzas per year)

34
Real vs. Nominal Variables

• Nominal variables are measured in terms of some
  currency (e.g. your nominal income is 70,000 per
  year)
• Real (inflation adjusted) variables are measured
  in terms of some commodity (e.g. if pizzas cost
  10/pizza your real income is 7,000 pizzas per
  year)
• Real Nominal/Price ( 7000 70,000/10 )

35
Empirical Questions

• What exactly are you trying to measure? Is your
  variable consistent with what youre trying to
  measure?
• How is your variable measured?
• Is your variable in real or nominal terms?
• Is your variable seasonally adjusted?

36
Example Seasonality
37
Components of Economics Data

• Economic data series are generally believed to
  have four main components

38
Components of Economics Data

• Economic data series are generally believed to
  have four main components
• Trend (many years)

39
Components of Economics Data

• Economic data series are generally believed to
  have four main components
• Trend (many years)
• Business Cycle (1-2 yrs)

40
Components of Economics Data

• Economic data series are generally believed to
  have four main components
• Trend (many years)
• Business Cycle (1-2 yrs)
• Seasonal ( lt 1 yr)

41
Components of Economics Data

• Economic data series are generally believed to
  have four main components
• Trend (many years)
• Business Cycle (1-2 yrs)
• Seasonal ( lt 1 yr)
• Noise (very short term)

42
Components of Economics Data

• Economic data series are generally believed to
  have four main components
• Trend (many years)
• Business Cycle (1-2 yrs)
• Seasonal ( lt 1 yr)
• Noise (very short term)
• Typically, we are not interested in the seasonal
  component, so we remove it.

43
Seasonally Adjusted Retail Sales
44
Empirical Questions

• What exactly are you trying to measure? Is your
  variable consistent with what youre trying to
  measure?
• How is your variable measured?
• Is your variable in real or nominal terms?
• Is your variable seasonally adjusted?
• Is your variable annualized?

45
Example Annualizing

• A 90-day T-Bill currently sells for 99.80 per
  100 of face value. This implies a 90-Day return
  of around .2

46
Example Annualizing

• A 90-day T-Bill currently sells for 99.80 per
  100 of face value. This implies a 90-Day return
  of around .2
• A 5 year STRIP currently sells for around 90.25
  per 100 of face value. This implies a return of
  around 10.8

47
Example Annualizing

• A 90-day T-Bill currently sells for 99.80 per
  100 of face value. This implies a 90-Day return
  of around .2
• A 5 year STRIP currently sells for around 90.25
  per 100 of face value. This implies a return of
  around 10.8
• How can we compare these two rates of return?

48
Example Annualizing

• Annualizing converts any data series to a common
  time frame (1 year)

49
Example Annualizing

• Annualizing converts any data series to a common
  time frame (1 year)
• Assuming that the 90 day interest rate stays
  constant at .2, the annual return to 90 day
  T-bills would be (1.002)(1.002)(1.002)(1.002)
  1.008 .8

50
Example Annualizing

• Annualizing converts any data series to a common
  time frame (1 year)
• Assuming that the 90 day interest rate stays
  constant at .2, the annual return to 90 day
  T-bills would be (1.002)(1.002)(1.002)(1.002)
  1.008 .8
• What would your annual return need to be to
  receive a (compounded) 5 year return of 10.8
• (1x)(1x)(1x)(1x)(1x) 1.108
• x 1.02 (2)

51
Example Annualizing

• Annualizing converts any data series to a common
  time frame (1 year)
• Assuming that the 90 day interest rate stays
  constant at .2, the annual return to 90 day
  T-bills would be (1.002)(1.002)(1.002)(1.002)
  1.008 .8
• What would your annual return need to be to
  receive a (compounded) 5 year return of 10.8
• (1x)(1x)(1x)(1x)(1x) 1.108
• x 1.02 (2)
• These two annualized rates can now be compared