Smog Reduction

1
Smog Reductions Impact on California County
Growth

• Study looks at the relationship between changes
  in environmental quality and population change.
• San Bernardino and Riverside counties suffer from
  highest ozone levels in the country.
• Ozone a strong irritant that can cause
  constriction of the airways, forcing the
  respiratory system to work harder to provide
  oxygen. For healthy people it makes breathing
  more difficult.but may pose a worse threat to
  those who are already suffering from respiratory
  diseases such as asthma..

2

• Due to vehicle and manufacturing regulations in
  California the number of high ozone days has
  drastically decreased in the counties San
  Bernardino had 40 fewer high ozone days in 1996
  compared to 1980.
• The papers thesis county quality of life
  increased in areas where ozone fell sharply and
  this has encouraged in-migration.

3

• The population of many California counties
  increased over time.
• The authors task is the link the timing of this
  growth to changes in environmental quality.
• This link would imply a relationship between
  environmental quality and migration.
• Typically a study wants to take local evidence
  (the sample) to say something generally about a
  population.

4

• Author first estimates equation to establish that
  San Bernardino and Riverside Counties growth has
  accelerated over period that pollution declined.
• Runs the regression model using data for all
  counties in California
• log(Popj,t1/Popj,t)? log(Popj,t)ßXjtU
• There are 58 counties in California.
• Author calculates logged change in population
  over two time periods 1969 to 1980 and 1980 to
  1994.
• So each county is observed twice the author
  stacks the data over the two times periods
• Pop is population in a given year and Xit is a
  series of dummy variables accounting for such
  factors as which time period the observation is
  in and whether the observation represents San
  Bernardino/Riverside County
• Why does author use only California counties?

5
(No Transcript)
6
Regression Model Results

• The coefficient for the log 1969 county
  population implies the inverse relationship
  between the population size of county in 1969 and
  subsequent growth over the two time periods.
• Larger counties in 1969 did not grow as fast in
  percentage terms as smaller.
• Significance?

7

• Coefficient for 1994 Calendar-Year dummy implies
  counties on average grew in population more
  slowly over the 1980-1994 period than in
  1969-1980 period.
• Los Angeles Region grew at faster pace than the
  remaining counties in the state.

8

• Growth in San Bernardino/Riverside relative to
  rest of the state
• Define
• X21 if dependent variable is an observation
  indicating growth over the 1980-1994 period 0
  otherwise
• X41 if county is San Bernardino or Riverside 0
  otherwise
• Model
• population growth .b4X4b6(X2X4)
• .058X4.421(X2X4)

9

• Model
• population growth .b4X4b6(X2X4)
• .058X4.421(X2X4)
• Growth in San Bernardino/Riverside relative to
  the other California Counties in 1969-1980
  period
• X20 and X41
• Over 1969-1980 San Bernardino/Riverside grew
  roughly 5.8 faster than the remaining counties
  in California

10

• Model
• population growth .b4X4b6(X2X4)
• .058X4.421(X2X4)
• Growth in San Bernardino/Riverside relative to
  the other California Counties in 1980-1994
  period
• X21 and X41
• Over 1980-1994 San Bernardino/Riverside grew
  roughly 47.9 faster than the remaining counties
  in California.

11

• The period of accelerated growth coincides with
  period of decreased pollution
• Regression model is more descriptive than one
  suggesting causation.
• Does the finding on the increased population
  growth prove the authors thesis?
• Could other factors have accounted for the
  growth?
• What role does the concept of statistical
  significance play in the authors thesis?

12

• Author more closely investigates relationship
  between environmental quality and population
  growth
• Runs regression estimating the determinants of
  California county growth only over 1980-1994
• Includes continuous variables that can impact
  county growth in regression, such as 1980 home
  price, 1980 percent hispanic
• One of the variables in the regression is the
  difference in the number of high ozone days in
  1980 and 1994.
• For example if a county had 40 high ozone days in
  1980 and 32 high ozone days in 1994 then the
  value of the variable for that county is -8
  indicating pollution fell in the county by that
  magnitude

13
(No Transcript)
14

• Log/linear Model
• Interpret coefficients in the two models
• Analyze significance of estimated effects
• R2?
• Why should author include other factors in
  regression model?
• Authors finding
• A county that experienced a 10-day reduction in
  high ozone days between 1980 and 1994 grew by 7.8
  percent more than a county whose ozone level
  remained unchanged.

15
(No Transcript)