Transcript No Slide Title

Economic Models for
Impact Assessment
Steven R. Miller
Senior Research Analyst
Center for Economic Analysis
Presented to the MI-SBTDC
Economic Models for Impact
Assessments
• An economic model is a simplification
of reality
• Models are based on mathematical
equations that provide a simplified
picture of employment and income
relationships
Economic Models for Impact
Assessments, continued
• Economic impact models are used to
assess the impacts of shocks to the existing
economic structure
• Shocks take many forms
– Increase in export demand for a locally produced
good or service
– Increase in capital from outside the region (SBA
loans and grants)
– Changes in federal, state, and local tax codes
– New social structures like an arena
– Reallocation of resources from one activity to
another
Economic Models for Impact
Assessments, continued
• The accuracy of the impact assessment
depends on:
– The appropriateness of the model assumptions
– The accuracy of the shock supplied to the model
– The user’s understanding of the model structure
in interpreting the results
Appropriateness of the Model Assumptions
• Consider two models of the labor market
using different assumptions about wage
rates.
– Both models are valid, but the concept of validity
is a moving target
Wages are assumed fixed
Wages are assumed flexible
Wages
Wages
S
S
D
0
A
B
C
D`
Employment
D
0
A
B
C
D`
Employment
Appropriateness of the Model Assumptions
• The fixed wage model produces a greater
increase in labor for a given increase in
demand
• The fixed wage model is most appropriate
for
– Small shocks to the economy
– Economies with high unemployment
• The flexible wage model is most appropriate
for
– Large shocks to the economy
– Economies with low unemployment
Types of Impact Assessment
Models
Custom or Off-the-shelf?
• Custom models generally
– Better represents the local economy
– Are extremely expensive to acquire
• Off-the-shelf
– Standardized structure across all regions
– Less expensive to acquire
Off-the-shelf Models
• All off-the-self models share common
characteristics
– Use nationally standardized and reported data
– Rely on a core Input-Output model to capture
relationships across industries
– Regional detail for models are generally derived
by adjusting national data to fit the local
economy
• Hence, off-the-self models are fairly
comparable.
The Input-Output Table
• Input-Output table estimates what each
industry buys from all other industries within
the region to make a dollar’s worth of
output.
• When one industry increases production, it
increases its purchases of inputs from other
industries.
• Hence a $1 increase in production will
generally lead to more than $1 increase in
production in all industries
The Input-Output Table: Example
• Example
– Steel industry uses coal to fire its blast furnaces
and to mix with iron ore
– Coal industry uses steel in its equipment and to
brace tunnels
• When steel is sold, demand for coal increases
• When coal is sold, demand for steel increases
The Input-Output Table:
Example
• Imagine that
– $100 of steel requires $10 of coal
– $100 of coal requires $5 of steel
• Imagine that both industries are found
locally
• Consider the local impacts if a new
customer buys $100 worth of steel
Direct Impact
Interindustry Purchases
Interindustry Purchases
Interindustry Purchases
$ 100.00
$ 10.00
$
0.50
$
0.05
$ 110.55
Steel
Coal
Steel
Coal
Total Impact
The Input-Output Table:
Example
• Hence a $100 purchase has led to more
than $100 worth of local production
• The ratio of Total Impact to Direct Impact is
known as the multiplier and is always
greater than 1
• In our example:
Total Impact/Direct Impact = 110.55/100 = 1.1055
• This multiplier is the same if the increase in
demand for steel is $100, $200, or $1 million
• Simply take the dollar value and multiply it by
the multiplier to get the total impact
The Input-Output Table,
Continued
• This example demonstrates the working of
the core of all off-the-shelf models
• Rather than limiting the relationships to two
industries, these models track the local input
requirements of all local industries
– RIMS, IMPLAN, and EconomicImpact report the
full spectrum of federally reported industries
– REMI aggregates industries into 169 categories
The Input-Output Table,
Continued
• In a similar fashion household wages and
expenditures are recorded. In the steel/coal
example,
– both the steel and the coal industries hire
additional workers to produce the added output
– These additional workers spend part of their
wages on goods and services provided locally
– Benefiting firms will similarly hire additional
workers to accommodate the increased demand.
Off-the-shelf Models,
Continued
• Several options exist for off-the-shelf
models.
– BEA RIMS II (Bureau of Economic Analysis)
– Impact and Planning (Minnesota IMPLAN
Group)
– Policy Insight (REMI, Inc.)
– EconomicImpact (EMSI)
Back to Off-the-Shelf Models
RIMS II
• Provides only the Multipliers for 473 industries
• Useful in building custom models around the inputoutput table
• Inexpensive (about $225 per district)
• Static, making no room for modification of the
underlying data
• Data is updated every five-years
Back to Off-the-Shelf Models
IMPLAN Pro.
• Provides the modeling environment used to
produce the RIMS II multipliers for 509 industries
• Dynamic in that it allows the user to provide
informed modifications to the underlying data
• Inexpensive ($2000 per year)
• Data is updated annually
• Expands on RIMS to include flows to and from
government sectors
• STRICTLY DEMAND DRIVEN
Back to Off-the-Shelf Models
Policy Insight
• Expands on the IMPLAN methodology with greater
model coverage for 169 industries
• Dynamic in that it allows the user to provide
informed modifications to the underlying data
• Expensive (varies based on number of industry
components and geographic area for model)
• Relaxes many of the model assumptions fund in
IMPLAN
• PROVIDES BOTH SUPPLY-SIDE AND DEMANDSIDE IMPACTS
• Provides forecasts
Back to Off-the-Shelf Models
EconomicImpact EMSI (Limited information
•
•
•
•
•
on)
This model seems to be built to identify human
resource needs (training, occupations, etc.)
Provides the modeling environment used to
produce the RIMS II multipliers
Dynamic in that it allows the user to provide
informed modifications to the underlying data
before calculating the multipliers
Web-based
Provides forecasts
Comparison Tableau
RIMS II
IMPLAN
REMI
Industries
Updates
Periodicity
Structure
473
Every 5-years
Annual
Input-Output
169
Yearly
Annual
Hybrid Model
Employment Impacts
Income Impacts
Output Impacts
Demographic/Household
Impacts
Includes Fiscal Impacts
Provide Tax Policy
Impacts
Provides Baseline
Forecasts
Modifiable
Provides Occupational
Impacts
Demographic Profiles
No
Yes
Yes
No
509
Yearly
Annual
Input-Output
(SAM)
Yes
Yes
Yes
No
Economic
Impact
500
Every 6-months
Annual
Input-Output?
Yes
Yes
Yes
Yes
Yes
??
??
No
No
Indirectly
Yes
Indirectly
Yes
Yes
No
Indirectly
No
No
Yes
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
None
No
By Age, By
Race, By Gender
Yes
By Age Group
Facilitates Cluster
Analysis
Cost to Acquire
(excludes analysis)
By Income
Group
Yes
$225 per district
per 5-years
$2,000 per year
for all districts
Variable (pricey)
Variable
(Not outrageous)
Yes
In Output
Direct Impact
Input-Output
Increases demand for locally produced
inputs
Suppliers to auto manufacturing increase
demand for locally produced inputs into
their production process
Increases in wages lead to increased demand
for locally produced goods and services at
the retail level
SAM Feedback
Increased wages and profits leads to
increased income tax collection
Increased income leads to increased sales
tax revenue
Hybrid Component
Increase in the demand for workers
Increases the wage rate
causes the average local price to increase
relative to the nation
causes local production costs to increase
making locally produced goods and
services less competitive to the nation
causes increase in demand for taxable
productive capital
Increase in wages and jobs increases labor
in-migration
Increase in-migration rates lead to
demographic shifts and increased
government expenses and revenue
Output Multiplier
Employment Multiplier
RIMS
$200 Mil
IMPLAN
$200 Mil
REMI
$200 Mil
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes1
Yes
No
No
Yes
No
No
Yes
No
No
Yes
No
No
Yes
No
No
Yes
No
No
Yes
1.826498
4.381753
1.474
2.909
• For further information, contact
Steven R. Miller
88 Agriculture Hall
Michigan State University
East Lansing, MI 48824
(517) 355-2153
[email protected]