Download Economic Research and Publication

Ec 596
Economic Research and Publication
Portland State University
April 15, 2010
Office of Economic Analysis
Josh Lehner
Overview
• My Research Project
– Origination, Plan, Analysis
•
•
•
•
•
•
•
Presentation and Writing
Publication of Results
Advisor’s Role
What I Do
Future Research
Possible Project Ideas
State of the State (pending)
2
My Research Project
• Retail Sales Tax Discrepancy
• Washington – Oregon; Portland – Vancouver
• Intuitively it should have an affect
• Empirically does it?
• If so, what are the consequences?
• Panel Data
• FE, Spatial
• Calculated fiscal impacts
3
Project Origination
• Previous employer – CREDC
• How Vancouver fits into MSA
• PSU courses
• Regional/Urban Economics is interesting
• Highly marketable
• Original idea: price point survey
• Literature Review and Advisor changed my mind
• Broader, more macro level analysis
4
Plan and Design
• Idea
• Literature Review
• Theoretical and Empirical
– Methodologies
• Find citation/reference but no paper – email author
• Read paper, if useful, write proper citation immediately
– Download paper
• Short summaries combined into one larger file
• Don’t be intimidated by fancy math or what appears to be
fancy math
5
Plan and Design (cont)
• Modify Idea as needed
• Specify model/equation with all possible
variables (easier to downsize)
• Core and Choice Variables
• Identify needs, sources, descriptions
• Build Data Set
• Univariate, multivariate
• Further methodological research, learning (Hausman)
• Statistical program learning
6
Misc. Data and Model
• Panel data
• Need a set of items, data over time
– Countries, States, Counties, Firms
– Balanced panel best (especially for spatial)
– May be able to compute missing values – it depends
• Time Series or Cross Sectional
• Index, Forecasting, VAR
• Don’t reinvent the wheel
• Literature review helps formulate ideas
7
Summary of Project
8
Reexamining the Border Tax Effect:
A Case Study of Washington State
Rossitza Wooster & Josh Lehner
Prepared for PNREC
May 19th, 2009
9
Overview
• What is the Border Tax Effect?
• Previous Estimates
• Model
• Empirical Results
• Conclusion
10
What is the “Border Tax
Effect”?
• Why it occurs:
• Neighboring jurisdictions with different tax structures
• Border residents purchase goods in lowest taxing
jurisdiction to avoid higher costs
• What it affects:
• Governments’ tax revenue
– Public goods and services
• Societal welfare
– Illegal tax evasion
• Other Issues
• Political “3rd rail” in Northwest
• Economic efficiencies (consumption tax)
11
Washington State
• No income tax
• Rely heavily on sales tax
• 20% of population lives in
border counties
• Most regressive tax structure
in nation (Institute of
Taxation and Economic
Policy, 2000)
7.5 – 9.5%
6.0%
0.0%
12
Is there a Border Tax Effect?
Washington-Oregon Border : Real Per Capita Retail Sales
$16,000
$14,000
$12,000
$10,000
Oregon
Counties
$8,000
Washington
Counties
$6,000
$4,000
$2,000
$0
1992
1997
2002
2007
• Washington’s border counties generate far less
retail sales than their peers in both Washington
and Oregon
13
Previous Estimations
• West Virginia – 1988
• 2 Previous Washington studies
– Lorrie Jo Brown (1990)
• 1975-1987 data
• Price elasticity -1.8 (SR), -2.4 (LR)
– John Beck (1992)
• 1984-1988 data
• Legislative Changes
• Price elasticity between -2 and -3.2
• Other Examples: State-level analysis, Event studies,
Alcohol taxes, Canada and E.U. (tax harmonization)
14
Data
•
•
•
•
39 Washington counties
15 years (1992-2006)
585 observations
9 final variables
• Tried many more
• 5,265 data points in Fixed Effects model
15
The Model: Variables
Variable
Type
Description
Expected Sign
Sales
Dependent
Real Per Capita Taxable Sales
NA
Inc
Standard
Real Per Capita Personal Income
+
Price
Standard
Home County Tax Rate Relative to
Neighboring County Tax Rate
-
Border
Standard
Binary Indicator For Border Counties
NA
Travel
Standard
Mileage Distance Between Counties,
Adjusted By Gasoline Index
+
Unemp
Control
County Specific Unemployment Rate
-
Youth
Control
Percentage of County 18 Years And
Younger
+
Elderly
Control
Percentage Of County 65 Years And
Older
-
RetailEst
Control
Retail Establishments Per 1,000
Population
+
16
Price Variable
price 
PH (1  t H )
PN (1  t N )
• Prices Assumed To Be Equal In Competitive
Market
• Measures Relative Price Of Goods
• Key Variable: Sales Tax Effect
• Expected Sign: • Expected Range: -2 to -11
• Brown: -2.4
• Beck: -2.0 to -3.2
17
Travel Variable
 GPIt 

Travel  Dist 
 GPI 2006 
• Distance Between County Seats
• Exceptions (Brown): Eastern Washington
• U.S. Energy Information Administration
• Controls For Cost Of Gasoline
• Expected Sign: +
18
Descriptive Statistics (Univariate)
Variable
Overall
Interior Counties
Border Counties
Mean
Std Dev
Mean
Std Dev
Mean
Std Dev
Sales***
11,612
4,161
12,843
4,179
9,413
3,090
Inc***
28,060
5,491
28,840
6,343
26,666
3,035
Price***
1.020
0.031
1.002
0.002
1.059
0.024
Travel***
67.424
52.519
95.245
45.231
17.742
13.310
7.377
2.479
7.425
2.215
7.292
2.895
Youth***
27.691
3.726
28.063
4.103
27.027
2.823
Elderly***
14.078
3.700
13.680
3.432
14.790
4.047
3.680
1.132
3.880
1.188
3.315
0.921
585
663
585
663
375
375
Unemp
RetailEst***
N
210
210
*** Difference-in-means test of interior versus border counties is significant at the 1% level
19
Model Specifications
• 4 Models: Fixed Effects & SAR, Semi-Log & Log
• 1992-2006
• Fixed Effects
sales it   0  1 ln priceit   2 ln( priceit * borderi )   3 ln incomeit
  4 ln travelit   5 X control   it
• Spatial Autocorrelation
Sales Measureit   0  1 ( Demand Variables )   W  Sales Measureit   it
20
Spatial Specification
Sales Measureit   0  1 ( Demand Variables )   W  Sales Measureit   it
Rho – coefficient, check significance level
W * SalesMeasure – spatially lagged dependent variable
W – weighing matrix, row standardized, symmetric, queen contiguity
SalesMeasure – per capita sales
Epsilon – normal error term
21
342,225 Data Points in Matrix
22
Model (1): Semi-Log Specification;
Dependent Variable = Salesit
FE
Model (2): Double-Log Specification;
Dependent Variable = ln(Sales)it
SAR
FE
SAR
Intercept
-117,268.77 ***
(18,622)
-110,203.90 ***
(17,438)
-3.094 *
(1.591)
-2.474
(3.460)
Ln(Real, Per Capita Income)
10,048.16 ***
(1,003)
10,048.162 ***
(951)
0.856 ***
(0.090)
0.856 ***
(0.085)
Ln(County Relative Price)
71,078.39 **
(32,193)
71,078.41 **
(30,523)
6.730 ***
(2.182)
6.730 ***
(2.069)
Ln(CountyPrice*Border)
-104,389.46 **
(40,371)
-104,389.48 ***
(38,277)
-9.843 ***
(3.397)
-9.843 ***
(3.221)
Ln(Travel)
2,455.59 ***
(370)
181.04
(163)
0.249 ***
(0.035)
0.036 *
(0.019)
Ln(Unemployment Rate)
-83.902
(208)
-83.902
(197)
0.028
(0.021)
0.028
(0.020)
Ln(Youth Percentage)
8,678.53 ***
(2,748)
8,678.53 ***
(2,605)
1.160 ***
(0.265)
1.160 ***
(0.251)
Ln(Elderly Percentage)
-4,956.94 *
(2,775)
-4,956.94 *
(2,631)
-0.499 *
(0.262)
-0.499 **
(0.249)
Ln(Retail Establishments)
1,614.50 ***
(471)
1,614.50 ***
(447)
0.143 ***
(0.048)
0.143 ***
(0.046)
Spatially Weighted Retail Sales
(W ∙ Sales Measureit)
3.80e-08
(0.339)
5.54e-08
(0.339)
Time Dummies
Yes
Yes
Yes
Yes
Number of Observations
585
585
585
585
R2 / Log-Likelihood
0.3547
-4742.84
0.3193
694.08
23
Model (2): Double-Log Specification;
Dependent Variable = ln(Sales)it
FE
SAR
Intercept
-3.094 *
(1.591)
-2.474
(3.460)
Ln(Real, Per Capita Income)
0.856 ***
(0.090)
0.856 ***
(0.085)
Ln(County Relative Price)
6.730 ***
(2.182)
6.730 ***
(2.069)
Ln(CountyPrice*Border)
-9.843 ***
(3.397)
-9.843 ***
(3.221)
Ln(Travel)
0.249 ***
(0.035)
0.036 *
(0.019)
Ln(Unemployment Rate)
0.028
(0.021)
0.028
(0.020)
Ln(Youth Percentage)
1.160 ***
(0.265)
1.160 ***
(0.251)
Ln(Elderly Percentage)
-0.499 *
(0.262)
-0.499 **
(0.249)
Ln(Retail Establishments)
0.143 ***
(0.048)
0.143 ***
(0.046)
Spatially Weighted Retail Sales
(W ∙ Sales Measureit)
Price Elasticity
5.54e-08
(0.339)
Time Dummies
Yes
Yes
Number of Observations
585
585
R2 / Log-Likelihood
0.3193
694.08
• -3.113
• Good, Expected Result
– Significant at 1% level
• Previous Literature: -2 to -11
• Washington Studies:
– Brown, 1990: -1.8, -2.4
– Beck, 1992: -2 to -3.2
• Why Larger?
– Time Span Of Study
– Consumer Behavior
• Internet
• Bargain Shopping
24
Model (1): Semi-Log
Specification; Dependent
Variable = Salesit
FE
SAR
Intercept
-117.268 ***
(18,622)
-110,203 ***
(17,438)
Ln(Real, Per Capita
Income)
10048 ***
(1,003)
10,048 ***
(951)
Ln(County Relative Price)
71,078. **
(32,193)
71,078 **
(30,523)
Ln(CountyPrice*Border)
-104,389 **
(40,371)
-104,389 ***
(38,277)
Ln(Travel)
2,455 ***
(370)
181
(163)
Ln(Unemployment Rate)
-83.902
(208)
-83.902
(197)
Ln(Youth Percentage)
8,678 ***
(2,748)
8,678 ***
(2,605)
Ln(Elderly Percentage)
-4,956 *
(2,775)
-4,956 *
(2,631)
Ln(Retail Establishments)
1,614 ***
(471)
1,614 ***
(447)
Spatially Weighted Retail
Sales (W ∙ Sales Measureit)
Price Effect
• When Converting Units, A 1%
Increase In The Tax
Differential, Results In $333
Decrease In Per Capita Sales
• Following Brown and Beck
• Calculate “Lost” Sales And
Revenue Due To The Border
Tax Effect
3.80e-08
(0.339)
Time Dummies
Yes
Yes
Number of Observations
585
585
R2 / Log-Likelihood
0.3547
-4742.84
25
“Lost” Sales And Revenue
With Full Tax Harmonization
County
Asotin
Total Taxable Retail
Sales (2006)
Estimated Gain in
Taxable Retail Sales
Estimated State
Tax Revenue
Estimated Local
Tax Revenue
$183,624,442
$6,614761
$429,959
$33,074
Benton
$2,303,245,278
$404,105,660
$26,266,868
$4,849,268
Clark
$4,866,777,344
$967,864,073
$62,911,165
$5,807,184
$29,770,738
$10,431,561
$678,051
$146,042
Cowlitz
$1,337,394,181
$246,912,462
$16,049,310
$2,469,125
Garfield
$15,899,676
$5,291,463
$343,945
$52,915
Klickitat
$162,750,735
$46,204,117
$3,003,268
$231,021
Pacific
$195,060,498
$53,466,762
$3,475,340
$534,668
Pend Oreille
$89,831,028
$6,367,568
$413,892
$70,043
Skamania
$87,112,482
$24,539,598
$1,595,074
$122,698
$7,278,765,098
$281,190,667
$18,277,393
$4,217,860
Wahkiakum
$24,290,624
$9,865,906
$641,284
$98,659
Walla Walla
$718,942,577
$153,521,385
$9,978,890
$2,302,821
Whitman
$410,491,705
$23,420,572
$1,522,337
$304,467
$17,703,956,406
$2,239,796,555
$145,586,776
$21,239,844
Columbia
Spokane
Total
26
“Lost” Sales And Revenue
With 1% Reduction in Tax Differential
County
Total Taxable
Retail Sales
Estimated Gain in
Taxable Retail Sales
Estimated State Tax
Revenue
Estimated Local Tax
Revenue
Asotin
$183,624,442
$6,217,875
$404,162
$31,089
Benton
$2,303,245,278
$52,481,255
$3,411,282
$629,775
Clark
$4,866,777,344
$136,318,884
$8,860,727
$817,913
$29,770,738
$1,320,451
$85,829
$18,486
$1,337,394,181
$32,921,662
$2,139,908
$329,217
Garfield
$15,899,676
$705,528
$45,859
$7,055
Klickitat
$162,750,735
$6,600,588
$429,038
$33,003
Pacific
$195,060,498
$7,128,902
$463,379
$71,289
Pend Oreille
$89,831,028
$4,218,514
$274,203
$46,404
Skamania
$87,112,482
$3,505,657
$227,868
$17,528
$7,278,765,098
$149,031,053
$9,687,018
$2,235,466
Wahkiakum
$24,290,624
$1,315,454
$85,505
$13,155
Walla Walla
$718,942,577
$19,190,173
$1,247,361
$287,853
Whitman
$410,491,705
$13,792,115
$896,487
$179,297
$17,703,956,406
$434,748,110
$28,258,627
$4,717,530
Columbia
Cowlitz
Spokane
Total
27
Conclusion
• Purpose: Quantifiably Show Border Tax Effect
In Washington State
• -3.11 Price Elasticity
• $145.6 Million In Lost State Revenue
• $21.2 Million In Lost Local Revenue
• Future Work
• Internet Sales
• Washington Sales Tax Policy Changes
28
Presentation of Results
• Final Class(es) – all students
• Format – similar to what I just did, only better
– Condensed version of paper
•
•
•
•
•
Brief opening on issue
Review of existing literature and issues
Model specification
Empirical results
Conclusion and future work
• It’s been rumored that beverages calm the
nerves
29
Writing Style
• Practice
– Previous work experience
– Literature review (other academic readings)
• Professional, careful edits
– Proofread many, many times (take breaks - days)
• Proper citations (Harvard System)
– Can always change for specific journal if needed
• Know your audience
– Use proper jargon
• Elasticity, economies of scale, diminishing rate of return
30
Publication
• Long Process
• Still not physically published – 20 months and counting
– October issue a possibly (26 months)
• Profs have multiple projects running at various stages
• Finished Project June 2008
• Clean Graphics, Re-write/Edit
• Submit Paper Sept 2008
• Contemporary Economic Policy
• Upload to SSRN
• Wait
• Mid-December, Journal designates “major revision”
• Not very good, but not horrible enough to throw out
31
Publication (cont)
• Through holiday break and into Jan and Feb worked
evenings and weekends
• Respond to referee comments (3 referees, ~25 comments)
• Learned Spatial analysis
• Used county specific fixed effects
•
•
•
•
•
•
•
Resubmit end of February 2009
Wait
Wait some more
Acceptance June 2009, no revision required
Administrative issues (forms, copyright release)
Online “sneak peek” January 2010
Physical publication possibly in October 2010
32
Role of Advisor
• Help cultivate ideas (sounding board)
• Help sort literature that is relevant
• Discard tangential papers
• Don’t be afraid to ask questions
• Don’t need to understand everything in a paper at first
• Help understand data issues and methodology
• Also, statistical programs
• Help with edits, clarification of thoughts
• Provide honest critique, constructive criticism
33
What do I do?
• Oregon Office of Economic Analysis
• Forecast economy, tax revenues, population, prisons,
highway cost allocation study
• Oregon Economic Model
• Personal income, employment, housing
• Oregon Index of Leading Indicators
• Oregon Dollar Index
• Export Markets (International developments)
• Western U.S. states
34
My Future “Fun” Research
• Micro
– Firm/Item level retail sales
• Indexes for retail environment over time
• Identify market opportunities/vulnerabilities
• Market structure
• Macro
– Border Tax Effect
• U.S. Economic Census data
• Both sides of the river
• Do results hold up?
35
Economic Census
Washington-Oregon Border : Real Per Capita Retail Sales
$16,000
$14,000
$12,000
$10,000
Oregon
Counties
$8,000
Washington
Counties
$6,000
$4,000
$2,000
$0
1992
1997
2002
2007
• How do results compare?
• Simulations of tax changes
36
Research Projects – 2008
• Hedonic Models
– Urban Amenities
• Johnson Reid
– Energy Star
• Energy Trust of Oregon
• Survey – Discount Rate
• Voting Behavior
– Same-sex marriage
• Non-empirical
– Voting
– Corporations
37
Possible Research Projects
• Economic Activity Indexes
– Leading, Coincident, Lagging
• Regional
• Unemployment Rate Analysis
38
Economic Activity Indexes
• The Conference Board
• Methodology
• DSFM Program
– State Coincident Indexes (Philly Fed)
– Alan Clayton-Matthews
• http://users.rcn.com/alancm/dsfm/index.html
• Dallas Federal Reserve
– Texas Metro Indexes
• Oregon
– OILI – Oregon Office of Economic Analysis
– UO Index – Tim Duy, University of Oregon
39
Leading Indicators
10 of 11 Indicators are Positive
30%
Oregon Index of Leading Indicators
(Six-Month Annualized Percent Change, through February 2010)
10.0%
20%
6.7%
10%
3.3%
0%
0.0%
-10%
-3.3%
-20%
-6.7%
Recession in Oregon
-30%
-10.0%
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Leading Index (Left Axis)
Diffusion Index <50
Nonfarm Employment (Right Axis)
40
Oregon Dollar Index
Oregon Dollar index
190
175
160
145
130
115
100
85
70
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Oregon dollar index
Jan-02 Jan-03
Time
Jan-04
Jan-05
Jan-06
Jan-07
Jan-08
Jan-09
Jan-10
Fed weighted exchange index
41
Oregon Indexes
Oregon Indexes of Leading Indicators
(Six Month Annualized Percent Change, through February 2010)
30.0%
10.0%
20.0%
6.7%
10.0%
3.3%
0.0%
0.0%
OEA
UO Index
Jan-10
Jan-09
Jan-08
Jan-07
Jan-06
Jan-05
Jan-04
Jan-03
Jan-02
-10.0%
Jan-01
-30.0%
Jan-00
-6.7%
Jan-99
-20.0%
Jan-98
-3.3%
Jan-97
-10.0%
Total Nonfarm Employment (R)
42
Recession Probability
(through February 2009)
Oregon Probability of Recession
100%
1,800
90%
1,700
80%
70%
1,600
60%
50%
1,500
40%
1,400
30%
20%
1,300
Probability of Recession (L)
Program Source: Prof. Jeremy Piger, Univ of Oregon
Jan-10
Jan-09
Jan-08
Jan-07
Jan-06
Jan-05
Jan-04
Jan-03
Jan-02
Jan-01
Jan-00
Jan-99
Jan-98
Jan-97
Jan-96
Jan-95
Jan-94
Jan-93
Jan-92
Jan-91
0%
Jan-90
10%
1,200
Oregon Nonfarm Employment (R)
43
Pros and Cons
• Strengthen Time Series abilities
• Leading forecasts Coincident which forecasts Lagging
• Learn a new program (DSFM)
• Use a Gauss program designed by the Fed
• Rewrite for other platform(s)?
• Customize for region
• People would use it
• Publishable? Probably not
• May not involve a regression…
44
Unemployment Rate Analysis
• Oregon consistently has a high rate – Why?
–
–
–
–
Migration – “The young and the restless”
Unemployment Compensation
Seasonality in Employment
JOLT/BED data
• “Churn”
– Industry Mix
• Durable Manufacturing
– Wood Products, High Technology
– Exports
45
Unemployment Rates
14
35
U.S Unemp Rate
Oregon Unemp Rate
12
Oregon Rank (R)
30
10
25
8
20
6
15
4
10
2
5
0
0
1976
1980
1984
1988
1992
1996
2000
2004
2008
46
Unemployment Benefits
UI Benefits Paid (Jan 2005 - Mar 2010)
$350,000,000
Total
$300,000,000
Regular Program
$250,000,000
$200,000,000
$150,000,000
$100,000,000
$50,000,000
$0
Jan-05
Jul-05
Jan-06
Jul-06
Jan-07
Jul-07
Jan-08
Jul-08
Jan-09
Jul-09
Jan-10
• Generous benefits?
• Duration of the unemployment?
• Industry mix?
47
Employment Turnover
Employment "Churn" (1992 - 2009 Q2)
Percent of Private Employment either Gaining or Losing a Job
22%
U.S.
Oregon
20%
18%
16%
14%
12%
10%
1992Q1
1994Q1
1996Q1
1998Q1
2000Q1
2002Q1
2004Q1
2006Q1
2008Q1
48
Employment Turnover
Employment "Churn" (1992 - 2009 Q2)
22
21
20
19
18
17
16
15
14
13
12
1992Q1
California
Nevada
United States
1994Q1
Idaho
Oregon
Washington
1996Q1
1998Q1
2000Q1
2002Q1
2004Q1
2006Q1
2008Q1
49
Seasonality in Employment
1.100
Alaska
Idaho
Oregon
US
1.080
1.060
1.040
1.020
1.000
0.980
0.960
0.940
0.920
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
Jan-03
Jan-04
Jan-05
Jan-06
Jan-07
Jan-08
Jan-09
50
What Not To Do
51
Oregon Exports
Oregon Exports by Industry
Oregon's Total Exports
(1Q 1997 - 4Q 2009, current dollars)
6,000.0
5,000.0
20.0
Total Exports
(left scale)
10.0
3,000.0
0.0
(% change)
($ million)
4,000.0
2009 Q4
Total All
Industries
4,237.9
4,374.9
3.2%
Computer And
Electronic
Products
1,805.1
1,995.3
10.5%
Agricultural
Products
543.6
712.6
31.1%
Machinery,
Except Electrical
301.2
333.8
10.8%
Chemicals
420.9
289.0
-31.3%
Transportation
Equipment
245.4
211.7
-13.7%
79.7
112.5
41.1%
168.3
101.9
-39.5%
Food And Kindred
Products
94.2
108.2
14.9%
Wood Products
96.1
94.6
-1.5%
103.9
60.9
-41.4%
40.0
30.0
Y/Y %
Change
2008 Q4
-10.0
2,000.0
-20.0
1,000.0
Year-over-year
percent change
(right scale)
0.0
Q3 2009
Q1 2009
Q3 2008
Q1 2008
Q3 2007
Q1 2007
Q3 2006
Q1 2006
Q3 2005
Q1 2005
Q3 2004
Q1 2004
Q3 2003
Q1 2003
Q3 2002
Q1 2002
Q3 2001
Q1 2001
Q3 2000
Q1 2000
Q3 1999
Q1 1999
Q3 1998
Q1 1998
Q3 1997
Q1 1997
-30.0
Waste And Scrap
-40.0
Primary Metal
Manufacturing
Paper
52
Exports by Industry
Oregon Exports by Major Industry
(1Q 1997 - 4Q 2009, current dollars)
2,400
2,200
Computer And Electronic Products
Agricultural Products
2,000
Machinery, Except Electrical
Chemicals
1,800
Transportation Equipment
1,600
($ million)
1,400
1,200
1,000
800
600
400
200
0
Q3 2009
Q1 2009
Q3 2008
Q1 2008
Q3 2007
Q1 2007
Q3 2006
Q1 2006
Q3 2005
Q1 2005
Q3 2004
Q1 2004
Q3 2003
Q1 2003
Q3 2002
Q1 2002
Q3 2001
Q1 2001
Q3 2000
Q1 2000
Q3 1999
Q1 1999
Q3 1998
Q1 1998
Q3 1997
Q1 1997
Source: WISERTrade
53
Exports to China
Oregon Export Markets (1997 - 2009)
12.8%
20
All Other
China
18
8.7%
9.1%
16
19.9%
Volume ($ billion)
14
6.5%
2.7%
12
1.4%
10
0.8%
1.4%
6.8%
5.6%
2002
2003
7.1%
5.1%
8
6
4
2
0
1997
Source: WISERTrade
1998
1999
2000
2001
2004
2005
2006
2007
2008
2009
54
Pros and Cons
• Huge Panel Data Set (FE, RE, Spatial)
•
•
•
•
•
•
•
•
Socio/Economic Factors (age, sex, education, etc)
Industry Mix, Seasonality
Unemployment Benefits, Avg Duration
Public Expenditures (education, health, etc)
Extremely useful to policymakers, public, me
Large existing literature
Publishable? Possibly, depending upon model
Forecast? VAR is standard method
55
Other Research Possibilities
•
•
•
•
•
•
•
•
•
•
•
•
•
•
County data – tourism, voting
Airport statistics (Fuel, recessions, income growth, +/- airlines)
Seasonal factors
Migration – Drivers Licenses, IRS
Indexes – Oregonian, Home Prices
LAUS v CES v Benchmarking
Prisons, Crime
Tobacco Consumption
ARRA impacts?
Cluster Analysis – Athletic Apparel in Portland
Sporting events attendance demand model
Transfer fees, player salaries (inequality)
CEO “union”
International Trade – elasticities, products, cross-sectional data?
56
Contact Information
[email protected]
(503) 378-4052
www.oregon.gov/das/oea
oregoneconomicanalysis.wordpress.com
twitter.com/OR_EconAnalysis
57
State of the State
58
The U.S. Economy in Recovery
• Financial crises generally do not lead to V-shaped
recoveries
• What will take the place of fading fiscal stimulus and
the inventory cycle
• Consumers remain cautious, weakening the strength of
the recovery
• Employers are expected to begin hiring soon
• Tax revenues are down sharply, forcing tough
decisions for local, state and federal governments
Copyright © 2010 Global Insight, Inc.
59
Unemployment Rises as the
Economy Contracts
8
(Annual percent change, 2005 dollars)
(Percent)
11
6
10
4
9
2
8
0
7
-2
6
-4
5
-6
4
-8
3
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Real GDP Growth
Copyright © 2010 Global Insight, Inc.
Unemployment Rate
60
Employment Losses
(through March 2010)
U.S. Recession Employment Losses
1.0%
1953
1981
1973
1990
2001
0.0%
-2.0%
-3.0%
-4.0%
63
57
1953
1960
1973
1981
2001
Recovery
54
48
45
42
39
36
33
30
27
24
21
18
15
12
9
6
-7.0%
3
-6.0%
51
1948
1957
1969
1980
1990
2007
60
-5.0%
0
% from NBER Peak
-1.0%
Number of Months from NBER Peak
61
U.S. Economic Growth by
Sector
2007
2008
2009
2010
2011
Real GDP
2.1
0.4
-2.4
3.0
3.0
Consumption
2.6
-0.2
-0.6
2.4
2.7
Residential Investment
-18.5
-22.9
-20.5
0.8
27.5
Bus. Fixed Investment
6.2
1.6
-17.8
1.7
7.6
Federal Government
1.3
7.7
5.2
3.8
-2.5
State & Local Govt.
2.0
0.5
-0.2
-1.2
0.4
Exports
8.7
5.4
-9.6
11.9
7.8
Imports
2.0
-3.2
-13.9
10.2
7.9
(Percent change)
Copyright © 2010 Global Insight, Inc.
62
Other Key U.S. Indicators
(Percent change unless noted)
2007 2008 2009
2010 2011
Industrial Production
1.5
-2.2
-9.7
5.1
4.7
Payroll Employment
1.1
-0.6
-4.3
-0.5
1.8
Light Vehicle Sales (Millions)
16.1
13.2
10.3
11.8
13.8
Housing Starts (Millions)
1.34
0.90
0.55
0.67
1.19
Consumer Price Index
2.9
3.8
-0.3
1.9
2.0
Core Consumption Deflator
2.3
2.3
1.7
1.2
1.9
Federal Funds Rate (%)
5.0
1.9
0.2
0.2
1.7
10-Year Treasury Yield (%)
4.6
3.7
3.3
3.9
4.1
Copyright © 2010 Global Insight, Inc.
63
Food and Energy Prices Swing
Consumer Price Inflation
6
(Year-over-year percent change)
4
2
0
-2
-4
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
All-Urban CPI
Copyright © 2010 Global Insight, Inc.
Core CPI
Employment Cost Index
64
A Sharp Retreat in Oil and
Gas Prices
140
($/barrel, WTI)
($/million Btu, Henry Hub)
14
120
12
100
10
80
8
60
6
40
4
20
2
0
0
1998
2000
2002
2004
Crude Oil (Left scale)
Copyright © 2010 Global Insight, Inc.
2006
2008
2010
2012
Natural Gas (Right scale)
65
U.S. Dollar Recovered Briefly
(Real Trade-Weighted Dollar Index, 2005=1.0)
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
1976
1980
1984
1988
1992
Major Trading Partners
Copyright © 2010 Global Insight, Inc.
1996
2000
2004
2008
2012
Other Important Trading Partners
66
Risks to the Forecast
Pessimistic (20% Probability):
• Inventory cycle and stimulus fade, no sustained expansion
• Home prices, starts, sales, and construction fall more sharply
• A true double-dip recession in which Real GDP falls again
Optimistic (20% Probability):
• Financial rescue and fiscal stimulus plans gain traction
• Business investment and exports show more robust growth
• Housing and consumer markets rebound more quickly
Copyright © 2010 Global Insight, Inc.
67
Real GDP Growth in
Alternative Scenarios
(Percent change, annual rate)
8
6
4
2
0
-2
-4
-6
-8
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Baseline (60%)
Copyright © 2010 Global Insight, Inc.
Pessimistic (15%)
Optimistic (25%)
68
Bottom Line
• The technical recession in the U.S. ended during the summer, with
the unemployment rate toping out at 10.1% in the fourth quarter of
2009. The labor market will remain weak with unemployment
averaging 9.6% in 2010.
• Credit markets are slowly returning to pre-Lehman collapse days but
risk premiums are still present.
• The housing market will continue to remain fragile. Housing starts
marginally improved in the second half of 2009 off their historic
lows. Prices will decline into 2010.
• Inflation is not a threat today or even next year but the stage is set
for carefully executed exit strategies to avoid inflation in the future.
• Economic growth returned in the third quarter and the fourth quarter
saw strong growth, but it will remain below potential throughout
2010.
• With furlough days and reduced weekly hours, employment gains
will lag the recovery – another “jobless” recovery.
Copyright © 2009 Global Insight, Inc.
69
Office of Economic Analysis
Oregon
70
Recent Oregon Economy Facts
• 10.6% unemployment rate for March 2010 (Mar US rate is 9.7%) is up
from the latest lowest rate of 5.0% in April 2007 and down from the
highest rate of 11.6% in May and June 2009.
• 35th fastest job growth at -3.03% for all states for February 2010 over
February 2009.
• Total nonfarm employment dropped -5.7% year-over-year for the 4th
quarter of 2009. Job losses (S.A.) from February 2008 to December
2009 (up 1,000 in January). The last six months’ losses averaged 3,250
per month versus 10,017 per month over the first six months of 2009.
• -0.1% personal income growth for 4th quarter of 2009 over 4th quarter of
2008. Annualized 4th quarter 2009 growth at 3.8%.
• Oregon exports increased 3.2% in the 4th quarter compared to the same
period last year but finished 2009 down 23% over 2008. (Export growth
is positive Q/Q and is expected to follow the global economy)
71
End of Recession in Oregon?
• The recession in Oregon either ended this summer or is close to
ending
– the Oregon economy generally follows the US economy
– the Oregon Index of Leading Indicators (OILI) has been increasing
the past eight months (July - February)
– other economic indicators, the University of Oregon Leading Index
and the Philadelphia Federal Reserve State Coincident Index, are
generally following the trend of the OILI;
– the Oregon unemployment rate has stabilized since mid-2009
– the job loss rate in the state has greatly slowed since last April
– Average weekly work hours in manufacturing have recently
increased but still below expansion periods.
– signs of improving conditions at publicly traded companies in
Oregon show improvements in stock prices.
• Expecting a “jobless” recovery.
72
Oregon and ARRA
Oregon’s spending allotment is $3.9 billion, plus tax relief measures.
Through December 31st, $2.5 billion has been awarded with $1.7 billion
spent.
Employment
Education
Health & Human Services
Transportation
Public Safety
Natural Resources
Workforce
Community Services
Housing
Funds Awarded
Funds Expended
Energy
0
100
200
300
400
500
600
700
800
Millions
Source: www.oregon.gov/recovery
73
Labor Force Growth
(through February 2010)
Seasonally Adjusted (Index Jan 1990 = 100)
140
135
130
125
120
115
110
OR Labor Force
105
US Labor Force
OR Emp
100
US Emp
74
Jan-10
Jan-09
Jan-08
Jan-07
Jan-06
Jan-05
Jan-04
Jan-03
Jan-02
Jan-01
Jan-00
Jan-99
Jan-98
Jan-97
Jan-96
Jan-95
Jan-94
Jan-93
Jan-92
Jan-91
Jan-90
95
Unemployment Rate by Region, January 2010
(Not seasonally adjusted for counties)
Oregon: 11.8%
U. S.: 10.6%
(seasonally adjusted: 10.7%)
(seasonally adjusted: 9.7%)
10.6%
12.2%
11.7%
15.1%
12.5%
14.4%
Source: Oregon Employment Department
75
Office of Economic Analysis
Historical Comparison
Recession
1981-82
1980-82
U.S.
Oregon
1990-91
U.S.
2001
Oregon
U.S.
2008-?? *
Oregon
U.S.
Oregon
Employment
Loss (in 000s)
2,734.3
123.3
1,498.3
12.3
2,657.3
60.1
7,020.3
150.2
(2.99)
(11.50)
(1.37)
(0.97)
(2.01)
(3.69)
(5.09)
(8.64)
Peak-to-Trough
5 Qtrs
12 Qtrs
5 Qtrs
3 Qtrs
9 Qtrs
10 Qtrs
8 Qtrs
8 Qtrs
Return to Peak
8 Qtrs
28 Qtrs
10 Qtrs
5 Qtrs
15 Qtrs
16 Qtrs
20 Qtrs
24 Qtrs
% Change
Duration
* Estimates based on Global Insight and OEA forecasts
76
Historical Comparison
(through February 2010)
Oregon Employment Loss by Recession
1973
% Job Loss from Peak Employment
0%
1969 1960 1948
1953
1990
1980
2001
-2%
-4%
-6%
-8%
1948
1960
1980
2007
-10%
1953
1969
1990
Recovery
1957
1973
2001
-12%
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
Number of Months from Employment Peak
77
Oregon’s Lost Decade?
1,900,000
2008 Q1
1,738,243
1,800,000
2000 Q4
1,627,407
1,700,000
1,600,000
2000 Q1
1,605,857
2010 Q4
1,602,375
1,500,000
1,400,000
2003 Q2
1,567,494
2010 Q1
1,588,089
1,300,000
1,200,000
1990Q1 1992Q1 1994Q1 1996Q1 1998Q1 2000Q1 2002Q1 2004Q1 2006Q1 2008Q1 2010Q1 2012Q1 2014Q1
78
Signs of Life
(1st Quarter 2010)
10%
Newly Expanding
Expanding
Educational Services
Wholesale Trade Retail Trade Leisure & Hospitality
5%
Prof. & Bus. Services
Metals & Machinery
Health Services
Nat. Resources
Information
Public Education
0%
T/W/Util
Govt excl Education
Food
Qtr-to-qtr %change
Finance
Electronics
-5%
Wood Products
-10%
-15%
-20%
Trans. Equipment
Construction
Contracting
-25%
-20%
Newly Slowing
-15%
-10%
-5%
Year-over-year % change
0%
5%
10%
79
Initial Claims
Oregon
Unemployment Benefit Initial Claims (1st week 2002 - Mar 27, 2010)
22,000
20,000
18,000
16,000
Raw
14,000
12,000
10,000
8,000
6,000
4-Week Moving Average
4,000
2,000
0
2002
2003
2004
2005
2006
2007
2008
2009
2010
80
Initial Claims
(through February 2010)
Initial Claims per 1,000 Labor Force (SA)
7
Oregon
U.S.
6
5
4
3
2
1
0
1987M01 1989M01 1991M01 1993M01 1995M01 1997M01 1999M01 2001M01 2003M01 2005M01 2007M01 2009M01
81
Housing Starts: Oregon & U.S.
Housing Starts Index
(1973-2007 Average = 100)
200
U.S. (1.55 million)
Oregon (21,200)
Oregon (December GI)
Previous Oregon Forecast
150
100
Office of Economic Analysis
2015Q1
2013Q1
2011Q1
2009Q1
2007Q1
2005Q1
2003Q1
2001Q1
1999Q1
1997Q1
1995Q1
1993Q1
1991Q1
1989Q1
1987Q1
1985Q1
1983Q1
1981Q1
1979Q1
1977Q1
1975Q1
0
1973Q1
50
82
Oregon Housing Permits
(through February 2010)
Oregon Housing Permits (Monthly, SA 3 MMA)
3,500
Total Permits
Single Family
3,000
2,500
2,000
1,500
1,000
Jan-10
Jan-09
Jan-08
Jan-07
Jan-06
Jan-05
Jan-04
Jan-03
Jan-02
Jan-01
Jan-00
Jan-99
Jan-98
Jan-97
Jan-96
Jan-95
Jan-94
Jan-93
Jan-92
Jan-91
0
Jan-90
500
83
Mortgage Loans
• 6.88 percent of all loans past due (4th Quarter, 2009)
– 2001 peak 3.72%, rising since early 2007
– Oregon ranks 8th best nationally (US average is 10.44%)
• 2.98 percent of all loans in foreclosure (4th Quarter,
2009)
– Higher than 2002 (1.34%) and rising since late 2006
– Oregon ranks 25th best nationally (US average is 4.58%)
• Combined 9.86 percent ranks 10th best nationally
– US average is 15.02 percent
Source: Mortgage Broker’s Association
84
Oregon was Late to the Run Up in
Prices (Jan 2005 - Dec 2009)
Housing Price Index
(12-month percentage changes)
50%
AZ
Oregon
Washington
Arizona
California
Idaho
Nevada
40%
30%
ID
20%
10%
NV
OR
0%
CA
-10%
WA
-20%
-30%
-40%
JUN-05
DEC-05
JUN-06
DEC-06
JUN-07
DEC-07
JUN-08
DEC-08
JUN-09
DEC-09
Source: LoanPerformance
85
Selected State and US House
Price Appreciations
Annual Percentage Change in FHFA House Price
Indexes through 2009 Q4
30
California
Washington
25
20
15
Oregon
10
US
5
0
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
-5
-10
-15
-20
-25
Source: Federal Housing Finance Agency
86
Selected Oregon MSA House
Price Appreciations
40
Annual Percentage Change in FHFA MSA House
Price Indexes through 2009 Q4
Bend
35
30
Medford
25
20
15
10
5
Eugene-Springfield
PDX-Vanc-Bevrtn
Salem
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
(5)
(10)
(15)
(20)
Source: Federal Housing Finance Agency
87
Risks to the Forecast…
Downside
• Double Dip in Housing
• Higher Oil Prices
• Premature Policy Tightening
• After Shocks from the Financial Crisis
• China Bubble?
Upside
• V-Shaped Recovery in Other Parts of the World
• Quicker, Stronger Release of Pent-up Demand
• Stronger Growth in Total Factor Productivity
Upside – Downside?
• Passage of Measures 66 and 67
• Health Care Reform
88
Bottom Line for the Oregon Economy
• The “technical” recession in Oregon ended late
summer or late 2009. Expect a “jobless” recovery.
• Job losses will continue into the first quarter of 2010,
with only mild job growth the rest of the year.
• Housing prices may still decline into 2010 but looking
more like a bottom has been reached in housing
permits.
• Housing will not lead during the recovery. First
sectors likely to come back: profession and business
services, health care services, computer and
electronic products, retail.
89
Contact Information
[email protected]
(503) 378-4052
www.oregon.gov/das/oea
oregoneconomicanalysis.wordpress.com
twitter.com/OR_EconAnalysis
90