=round(#,0)
Feb 24, 2016
=round(#,0)
Figure 1. California Manicurists, 1987-2002
𝑁𝑚=𝛿0+𝛿1 ∙𝑉𝑚+𝛿2 ∙𝑇+𝜇𝑚
Empirical Specification
𝑚=1…34metro areas
0 5000 10000 15000 20000 250000100020003000400050006000700080009000Los Angeles
CA Manicurists, 2002
Non-Viet
Manicuri
sts (#)
Viet Manicurists (#)
Non-Viet Manicurists (#)
Viet Manicurists (#) Time Trend…T=0,…,15
𝛿1= 0.74 5(25.39 )
Do Vietnamese manicurists displace native manicurists?
𝑁𝑚=𝛿0+𝛿1 ∙𝑉𝑚+𝛿2 ∙𝑇+𝜇𝑚
Empirical Specification
𝑚=1…34metro areasNon-Viet Manicurists (# per 1000 residents)
Viet Manicurists (# per thousand residents)Time Trend…T=0,…,15
Do Vietnamese manicurists displace native manicurists?
Non-Vietnamese Manicurists per 1,000 ResidentsOLS (2)
Viet. Manicurists per 1,000 Residents -0.388***(11.40)
Time Trend -0.008**(2.25)
Constant 1.346***(53.92)
Observations 544
R-squared 0.346
Absolute value of t-statistics in parentheses
* significant at 10%; ** significant at 5%; *** significant at 1%
Table 1. Explaining the Number of Non-Vietnamese Manicurists using WLS
where
Assume
Empirical Specification
True Model
𝑁𝑚=𝛿0+𝛿1 ∙𝑉𝑚+𝛿2 ∙𝑇+𝜇𝑚
Unobserved demand shocks that vary across cities, e.g., increased demand by “tweens” and their parents in Contra Costa, according to a couple of newspaper articles.
𝐸 (�̂�1|𝑉𝑚)=𝛿1+𝛿2 ∙𝑐𝑜𝑣 (𝐷𝑚 ,𝑉𝑚)𝑣𝑎𝑟 (𝑉𝑚)
Increases in demand increase the number of native manicurists.
¿ ¿
Vietnamese manicurists are drawn to cities with positive demand shocks and away from those with negative shocks.
(− )Displacement effect Upward bias: the expected value of the estimated displacement effect is less negative than the true displacement effect.
¿
Instrumental Variable (IV) Estimation
Instrument relevance: Instrument exogeneity: z = historical size of Vietnamese enclave (VietEnclave)
instrumental variable
𝑉𝑚 (𝐷 ,𝑉𝐸 )=𝑉𝑚 (𝐷 )+𝑉𝑚 (𝑉𝐸 )
endogenous component, being part of the market for manicurists. exogenous component, being determined outside of the market for manicurists
Cov(VE, Cov(VE,
Hence, the instrument, , is likely to be relevant. Instrument purges of the component attributable to the demand shock due to some cities growing faster than others. Instrument exogeneity: 𝑐𝑜𝑣 (𝑉𝐸 ,𝜇)=𝑐𝑜𝑣 (𝑉𝐸 ,𝛿3 ∙𝐷𝑚 )≈0
The historical size of Vietnamese Enclaves (VE) is likely to be a valid instrument.
ivregress 2sls nonvietpc (vietpc= VietEnclave) T [aweight= pop]
Magic for now; explained after Thanksgiving Break
Non-Vietnamese Manicurists per 1,000 ResidentsOLS 2SLS or IV
Viet. Manicurists per 1,000 Residents -0.388*** -0.703***(11.40)
(11.88)Time Trend -0.008** 0.013***
(2.25) (2.66)
Constant 1.346*** 1.374***(53.92) (50.69)
Observations 544 544
R-squared 0.346 0.242
Absolute value of t-statistics in parentheses
* significant at 10%; ** significant at 5%; *** significant at 1%
Table 1. Explaining the Number of Non-Vietnamese Manicurists using WLS
Displace versus replacement
0.0 0.5 1.0 1.5 2.0 2.50.00.51.01.52.02.53.0
Non-Vietn
amese M
anicurists
(# per th
ou resid
ents)
Vietnamese Manicurists (# per thou residents)
Oxnard, CA198919901991
1987 1989 1991 1993 1995 1997 1999 2001 20030.0
0.5
1.0
1.5
2.0
2.5
Explaining City Fixed Effects
San DiegoRiverside
San Jose (Silicon Valley)
Non
-Vie
t Man
icur
ists
(# p
er 1
,000
res
iden
ts)
3. What is the return to learning English for immigrants?
where Mexican-born workers in 1980
Assume
𝑙𝑛𝑤 𝑖=𝛽0+𝛽1 ∙𝐸𝑛𝑔𝑖+𝜇𝑖
where 𝜇𝑖=𝛽2 ∙ 𝐴𝑖+𝜀𝑖
Empirical Specification
True Model
Instrumental Variables
Questions
1. What is the price elasticity of demand for cremations?
𝐶𝑃𝐼 2011
𝐶𝑃𝐼1991=224.9
136.2∙35.95=$ 59.36
1982−84=100
Instrumental Variables
Questions
1. What is the price elasticity of demand for cremations?
where
Assume
𝐶𝑅𝑚=𝛽0+𝛽1 ∙𝑃𝐶𝑚+𝜇𝑚
Empirical Specification
True Model
𝐸 ( �̂�1|𝑃𝐶𝑚)=𝛽1+𝛽2 ∙𝑐𝑜𝑣 (𝑃𝐶𝑚 ,𝑃𝐵𝑚)
𝑣𝑎𝑟 (𝑃𝐶𝑚)
Instrumental Variable
Instrument relevance:
Instrument exogeneity:
Example 15.3: z=cig tax, x=packs smoked suggests using R2E laws