The CNMI’s Econometric Model

The CNMI’s Econometric Model

Presented By:The Marianas EB5 Regional Center

For The:CNMI Department of Commerce

Paid By:The US Department of Interior

Office of Insular Affairs

Model BackgroundWho, What, Why & How

Background• The CNMI’s Economic Modeling

• Outdated and inconsistent with current economic realties• Doesn’t reflect impact of loss of manufacturing & tourism changes

• Current Economic Measures • 2007 & 2010 Gross Domestic Products

• Produced by the US Bureau of Economic Analysis• Provides a true measure of economic health

• Estimates are hindered by data gaps and the need to extrapolate data made available

• Economic Indicators• State-level collection of economic markers which provide a sense of economic condition• Not consistently collected and reported

• Consumer Price Index• State-level collection of prices in specific “market baskets” which gauge inflationary changes to

general consumer commodities• Historically only collected on Saipan, which didn’t reflect nature of inter-island price fluctuations

for the specific time collected• Recently developed Tinian & Rota CPI’s to account for actual changes in those specific economies

Need for a Model• To forecast impacts of specific projects, proposals and/or policy decisions• Useful in understanding multiplier impacts of specific economic

activity• Example: How many jobs a specific development will create (Direct,

Indirect & Induced)• Useful in forecasting impacts of policy implementation specific to the

economic condition• Example: How will changes to tax laws or structure impact revenues

• Useful in the application/evaluation of economic development programs• The CNMI’s Qualifying Certificate Program • Free Trade Zone Incentives

CNMI Steps

• CNMI Department of Commerce• Primary government entity for reporting economic condition through

its Central Statistics Division• Applied for a Technical Assistance Grant from the US Department of

Interior, Office of Insular Affairs• Placed RFP12-DOC-064• Contract awarded to the Marianas EB5 Regional Center

• Economist: Dr. Michael K. Evans of Evans, Carroll & Associates• Contractor report to Governor Inos in June 2013

• Identified data gaps and discussed impact to the economic model

CNMI’s Economic Model• Model Summary

• Model was developed without specific time series data points (2003-2010)• Wages & employment: Impacts the wage/employment sections of the model• Attempts to extrapolate data utilizing from the BEA, Economic Census & CNMI

data, however too many inconsistencies existed to make extrapolation useful• Availability of 2011 GDP estimates from BEA

• Anticipated release of 2011 GDP estimates will include revisions to 2009 & 2010 data

• CNMI’s model will be recalculated once the revisions are made available• Tourist spending data

• MVA data provide tourist arrivals, however spending patterns per market are not available

• If data collection occurs specific to tourist spending, model can accommodate changes to include these data points

Model Elements

Equations & elements of the CNMI Model

Model FrameworkCategories Utilized In Model Framework

Consumption Gross Domestic ProductFixed Investments EmploymentExport of goods WagesExport of services Consumer Price IndexImport of goods Implicit GDP DeflatorImport of services Domestic personal disposable incomeFederal government consumption & investment

Foreign measures of personal income

Territorial government consumption & investment

Exogenous variables*Foreign variables for Japan, China & Korea and includes values of the currencies & real GDP for those countries.

Key variable: Number of VisitorsImpacts: Total Demand

Consumption Charts: Consumer Durables

-8

-4

0

4

8

12

100

110

120

130

140

150

160

2003 2004 2005 2006 2007 2008 2009 2010

Residual Actual Fitted

• Excludes spending patterns of Japanese visitors

• Represents shopping patters in which consumers purchase genuine products vs. counterfeit products in their home country

Consumption Charts: Consumer Non-Durables

• Consumption closely tied to current conditions

-20

-10

0

10

20180

200

220

240

260

280

2002 2003 2004 2005 2006 2007 2008 2009 2010

Residual Actual Fitted

Consumption Charts: Consumer Services

-12

-8

-4

0

4

8

280

320

360

400

440

480

2003 2004 2005 2006 2007 2008 2009 2010

Residual Actual Fitted

• When housing expenditures rise, purchases of fuel & household non-durables (utilities, etc.) rise

• Increases in non-durables may serve as the cause of housing expenditure increases.

Consumption Charts: Net Foreign Travel

• Equation developed to maintain consistency with BEA statistics

• Similar to export of services

-12

-8

-4

0

4

8

280

320

360

400

440

480

2003 2004 2005 2006 2007 2008 2009 2010

Residual Actual Fitted

Fixed Investment Equation

• Includes housing & capital spending• Population decline consistent with the need for housing• Capital spending primarily for tourism based investments with

the exit of the garment manufacturing industry• Visitor arrivals critical independent variable for this equation

Export Charts

-1.2

-0.8

-0.4

0.0

0.4

0.8 160

200

240

280

320

2002 2003 2004 2005 2006 2007 2008 2009 2010

Residual Actual Fitted

• Goods export is minimal and statistically insignificant

• Export of services is closely tied to tourism and visitor arrivals

• Data similar to Net Foreign Travel (Consumption)

Export: Visitor ArrivalsJapan

• Decline consistent with JAL pullout and fluctuation with Japan’s Real GDP

China/Korea

-.15

-.10

-.05

.00

.05

.10

.15

11.8

12.0

12.2

12.4

12.6

12.8

13.0

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Residual Actual Fitted

-20,000

-10,000

0

10,000

20,000

120,000

160,000

200,000

240,000

280,000

02 03 04 05 06 07 08 09 10 11 12

Residual Actual Fitted

Imports

• Total demand is difficult to measure due to scarcity of data

• Employee compensation is used to measure aggregate demand

• Exports of goods are more clearly linked to imports due to the need to import raw materials

-60

-40

-20

0

20

40

200

400

600

800

1,000

2002 2003 2004 2005 2006 2007 2008 2009 2010

Residual Actual Fitted

Government Consumption Territorial

• Federal consumption is statistically insignificant

• Wages are used as a proxy for aggregate demand

Manufacturing Compensation

• Manufacturing wages are closely tied to export of goods

• Decline consistent with garment manufacturing exit

Compensation for Government Workers, Territorial

• Boosts in visitors increases the need for local services, which accounts for the rise in 2009-2010

Private Sector Wage

• Excludes manufacturing & government wages.

• Usually separated into several sectors, but due to data gaps, CNMI’s model is unable to provide sector specific information

• CNMI’s model utilizes the CPI as a correlation value versus standard output measures

Implicit GDP Deflator

• Constructed to close the gap between current & constant dollar GDP

Consumer Price Index for CNMI

• 1% increase in minimum wage= 0.3% in CPI

Final Five Equations

• Not estimated stochastically• 2010 Employment in manufacturing ,

trade, hotels, and restaurants, all other private, and government to relevant components of GDP.

• EM = XGX * 39/17• ETR = (CDX + CNX) * 3495/313• HER = XSX * 5779/180• EOTH = (CSX + IPX) * 12220/331• EG = GSLX * 5277/300• E= EM + ETR + HER + EOTH + EG

Utilizing the Model to ForecastSimulation of a $10,000,000.00 injection into the CNMI economy

Forecast Simulation & Results• Considered a forecast despite having known variables (visitor arrivals), due to pending release of GDP account figures• Current GDP for 2011 & 2012 to rise from 4.1% change from 2010 (last

known results)• Constant GDP for 2011 & 2012 to rise from 0.9% in 2011 to 6.0% in

2012• Increase in GDP for 2011 & 2012 due to increase in visitor arrivals of

20,000

• Model Simulations• Both test an impact of $10,000,000.00 into the CNMI economy

• First scenario arrives at the $10,000,000.00 via an increase in visitor arrivals of 20,000

• Second scenario arrives at the $10,000,000.00 via an increase in capital spending

Economic Impact of an Increase 20,000 Visitors

• Assumptions of model• Average visitor spends roughly $500.00 x 20,000 visitors= $10,000,000.00

• Simulation Purpose• Simulation looks at the economic impact of an additional 20,000 visitors

and its impact to overall economic impact over a three year period• Results of Simulation

• Increase of three year increase of $12 Million in first year, $17 Million in second year & $20 Million in third year of a $10,000,000.00 “shock” into the CNMI economy• Multiplier for an increase of 20,000 visitors into the CNMI economy is 2.0 over

three years

Economic Impact of an Increase in Capital Spending of $10 Million

• Multiplier is 1.2 versus 2.0 (increase in visitor arrivals)• Capital spending multiplier is lower due to the need to import goods,

whereas tourism based increase is higher due to the amount of domestic spending. • Money is remitted out of the local economy in import expenditures

• The impact of tourism spending creates and sustains higher employment numbers vs. capital spending