Colby Economic

O U T L O O K

2 0 0 2

The Colby Economic Outlook is produced by the studentsenrolled in Economics 473, a senior seminar in economic

forecasting at Colby College under the direction ofAssociate Professor Michael Donihue. This year’s CEO takes a

look at economic conditions in the US and around the worldand presents our short-term outlook for the Maine economy.

Also featured in this year’s edition is a forecast for howthe citizens of Maine will vote in the 2004 Presidential

election and a case study covering alternative stockmarket investment strategies.

Donna Pitteri (co-editor)Adam BentkoverKevin BrunelleConor Cooper

Albert GoodmanVanessa HalecoJonathan Hierl

Chingiz Mammadov

The conclusions and analysis presented in the CEO represent the views of the authors and do not necessarilyrepresent the opinions or recommendations of the faculty and staff at Colby College.

Grete RødZachary Shull

Matthew TabasBenjamin Winston

Harrison WreschnerChristopher Zeien

Professor Michael Donihue

The Colby Economic Review is authored, edited, and published by:

Colby Economic Outlook, Volume XI, Number 1, December 6, 2002.

Inside the CEO

Troubling Events Hamper the Global Economy ............................................. 4A Snapshot of Inflation and Unemployment in the US ................................ 4-5The Fed Tries to Steer the Economy Toward Recovery .................................. 6Passive Professionals: A Case Study in Stock Market Investment ............. 7-9In Maine A Reason for Optimism ............................................................ 9-13

Demographics .............................................................................. 9-10Total Employment in Maine ................................................................. 10A Continuing Transformation for Maine’s Economy ............................... 11Personal Income .......................................................................... 11-12Retail Sales ....................................................................................... 12Tourism ........................................................................................ 12-13Manufacturing Workers’ Weekly Hours ................................................ 13Housing Permits ................................................................................. 13Forecast for the Maine Economy ........................................................ 13

The Colby Coincident Index of the Maine Economy ................................ 14-15Predicting Presidents: Forecasting Maine’s Choice in 2004 ....................... 15

Troubling Events Hamper the Global Economy

Recent world events have created challengingstruggles for the global economy and an un-certain environment for the economic forecastsin this year’s CEO. The disconcerting eventsof the past 18 months have left investors andcorporations in flux. Consumers and corpora-tions are becoming more focused on cash con-servation than investment as they worry aboutthe future.

In the United States consumers continue tokeep the economy afloat. The hangover fromthe collapse of the tech bubble combined withuncertainties around the world have dealt asevere blow to the stock market. Yet the ap-parently large negative wealth effect has notpushed the economy back into recession. Con-sumer confidence is low but spending thisholiday season has been strong. Despite ag-gressive intervention on the part of the Fed,resulting in historically low interest rates, com-panies are having a hard time getting a clearpicture of future earnings and hence newspending on capital goods remains sluggish.Manufacturing output has shrunk for the pastthree months showing that demand is still notas strong as in the past and inventories re-main high. State governments are dealing with

dramatic revenue shortfalls in Maine andaround the country and cuts in spending arelikely to be met by higher taxes as they struggleto meet constitutional requirements to bringtheir budgets into balance. Real estate hasbeen one of the strong sectors in the economyas increased demand for housing and low in-terest rates have spurred growth in real es-tate values and new home constructions. Add-ing to the uncertainty are fears of a develop-ing real estate bubble in some regions of thecountry reminiscent of the 1980s.

The Israeli-Palestinian conflict, the specter ofwar with Iraq and the battle against terrorismhave left the Middle East in turmoil and thestability of world oil prices in doubt. In Asia,China continues to emerge as an economicpower and the relationship it has with theUnited States has improved significantly sincethe terrorist attacks in the US of September11th. In Southeast Asia a string of terror at-tacks, most notably in Bali, has crushed thetourism industry in that region. And in NorthKorea the revelation that Pyongyang has beendeveloping nuclear arms in violation of a 1994treaty has filled the region with more tension.

A Snapshot of Inflation and Unemployment in the US

The levels of inflation and unemployment inthe United States are key gauges monitoredby economists and policy makers in Washing-ton. Despite recent criticisms that he mis-handled the build up and collapse of the ‘techbubble’ in the stock market, Federal ReserveChairman Greenspan has generally been cred-ited with a remarkable period of low inflation,steady prices, and sustained output growthprior to last year’s recession. Thus civilianunemployment is now the main factor analystswatch in this consumer-driven economy.

In this year’s CEO we forecast the civilian un-employment rate using a model designed to

capture its historical relationship to the gapbetween actual output growth and theeconomy’s ‘potential’ growth rate, as estimatedand forecast by the Congressional Budget Of-fice. According to this week’s report from theBureau of Labor Statistics, more jobs were lostthan created through the first 11 months ofthis year as the economy continues to strugglewith the effects of last year’s recession. Im-provements in employment generally lag be-hind turning points in output growth as firmswait to expand their work force until aggregatedemand is clearly on the rise. Our forecast isfor the unemployment rate to average just over5.7% this year – an increase of almost a full

4

percentage point over 2001; showing littleimprovement in 2003; and then falling toroughly 5.5% in 2004 and settling in at about5.3% by the end of the forecast horizon.

Our unemployment rate forecasts are used asinputs into another model to produce our out-look for overall price inflation in the USeconomy based on a modified Phillips Curvedesigned to capture an assumed policytradeoff between lower inflation and higher un-employment. Our model incorporates inflation-ary expectations on the part of economicagents and forecasts the rate of growth in thechain-weighted price index for the businesscomponent of Gross Domestic Product. Thisprovides a broad-based measure of overallprices in the economy and is favored by policymakers. In addition to the unemployment rate,our model also incorporates the rate of growthin worker productivity and the gap betweenimport price inflation and the rate of ‘core’ CPI

Inflation & Unemployment Forecasts

Variable Actual Forecasts2001 2002 2003 2004 2005

Unemployment Rate (%) 4.79 5.73 5.67 5.46 5.26Inflation* (%) 1.73 0.71 1.40 1.21 1.42Assumptions          Real GDP Growth (%) 0.25 2.51 2.77 3.43 3.43Output Gap** 2.90 -0.43 -0.20 -0.66 -0.07Productivity Growth (%) 1.08 4.70 2.63 2.50 2.50Import Price Gap*** -7.86 2.39 2.00 2.00 2.00* Nonfarm GDP price index** Potential GDP growth – Actual GDP growth*** Import Price Inflation – ‘Core’ CPI Inflation

-1

0

1

2

3

4

5

2

3

4

5

6

7

8

1992 1994 1996 1998 2000 2002 2004

Inflation & Unemployment

Inflation (left axis)

Unemployment Rate (right axis)

Forecast

(minus food and energy prices) inflation. Thisbroad measure of prices tends to run about0.5 to 1.0 percentage points below reportedCPI inflation.

We predict that prices for the business sectorof GDP will rise by just 0.73 percent this yearand rise by another 1.4% next year. Inflationis forecast to moderate slightly in 2004 be-fore settling in at 1.4% in 2005.

An interesting feature of our Phillips Curvemodel is an ability to infer, for policy makingpurposes, a value for the long-run rate of un-employment in the US – the so-called Non Ac-celerating Inflationary Rate of Unemployment,or NAIRU. Our estimate of the level of unem-ployment consistent with an economy growingat its potential rate of growth and with stableprices lies in the range of 5.75 to 6 percent,depending on the assumed rate of productiv-ity growth in the economy.

5

The Fed Tries to Steer the Economy Toward Recovery

1

2

3

4

5

6

7

2001 2002 2003 2004 2005

Actual

Percent

Forecast

Recession2001:I - III

Output Gap = 0By 2005:IV

Monetary Policy & the Fed Funds Rate

matically and the resulting wave of refinanc-ing this year has put more money in consum-ers’ pockets and helped prop up the economy.

Our forecasts for future interest rate move-ments are based on a reaction function – aneconometric model that tries to capture howFed policy makers ‘react’ to current economicevents in determining their target Fed Fundsrate. We forecast the Fed Funds rate using amultivariate regression that incorporates move-ments in the discount rate, inflation, and theoutput gap. We then use these forecasts pre-dict future movements in other short and me-dium-term interest rates – essentially tracingout the yield curve – that are important for ourmodel of the Maine economy. Given our as-sumptions of modest 4 th quarter GDP growth,no change in the discount rate, and stable in-flation, our reaction function correctly predicteda Fed Funds rate cut in the 4 th quarter of 2003,although not quite of the same magnitude thatactually occurred. Our outlook for the future isfor modest tightening in monetary policy overthe forecast horizon to around 4% by 2005.

In light of the turbulent times the US economyhas been experiencing, the Federal ReserveBoard of Governors has done remarkably wellat using monetary policy as a stabilization tool.Throughout the exceptional growth in the1990’s the Fed kept monetary policy fairlystable with an average Fed Funds rate of5.25%. Towards the end of 2000 there wassome tightening of monetary policy in an at-tempt to slow the economy down. The interestrate on Federal Funds reached a high of 6.5%in the third quarter of 2000. After the burst ofthe technology bubble, the economy began toexperience a slowdown. In order to try and pre-vent a prolonged recession, the Fed cut itstarget rate on interbank lending on 10 occa-sions during 2001. On November 6th of thisyear the Fed cut this key indicator of monetarypolicy again to its current level of 1.25%– thelowest nominal rate in over 40 years. As aresult, interest rates have tumbled throughoutthe economy with the prime lending rate fall-ing below 4.75% and the rate on passbooksavings and money market accounts to lessthan 2%. Mortgage rates have also fallen dra-

6

Passive Professionals: A Case Study in Stock Market InvestmentRecent stock market performance has beencharacterized by extreme volatility, both highsand lows of equity prices as a reaction to theburst of the “Technology Bubble” subsequentlyfollowed by a decline in Telecommunicationstocks, corrupt corporate governance, anduncertain investor confidence. Riddled withthe scars of the irrational returns of the latenineties, equity investors are currently tryingto pinpoint the market low. The Random WalkTheory (RWT) of equity prices, made popularby Burton Malkiel in his book, A Random WalkDown Wall Street, suggests that future stockprices are not dependent on past prices andin essence, are simply random. Modern Port-folio Theory (MPT), currently considered by aca-demics to be the optimal investment strategy,states consistent long-run returns can beachieved at fairly moderate risk through diver-sification. Combining his RWT with MPT,Malkiel suggests that market timing is not thekey to consistent, long-run returns. Rather, themost favorable means of investing is a pas-sive approach, holding a widely diversified port-folio, such as the S&P 500, for a lengthy time.

A now infamous test of this theory, popular-ized by The Wall Street Journal, “The DartboardPortfolio,” takes a group of randomly pickedstocks and compares their return to the “buyrated” analyst picks. The “dartboard” repre-sents the passive approach and the analyststhe active approach. For this year’s edition ofthe CEO we performed a similar experiment.We created a random portfolio by literally throw-ing darts at a recent copy of The Wall StreetJournal’s listing of current share prices. Thereturns from the Colby Random Portfolio (CRP)were then compared to five mutual fund port-folios, which act as a proxy for the analysts’active approach, and the performance of theS&P 500 – Malkiel’s proclaimed ‘superlative’investment. Some of the more well-knowncompanies represented in the CRP areLockhead Martin, Ralph Lauren, InternationalPaper, and Dave & Buster’s Restaurants. Themutual fund portfolios were selected from astratified sample of growth, value, small capi-talization and large capitalization stock fundsfrom the upper portion of the Lipper rankings.

 Portfolio

    CRP   S&P 500   FDEGX   FBGRX   FCONX   FMIEX   FMSPX

1 YEAR   -7.20%   -17.44%   -38.28%   -18.63%   -14.11%   -16.01%   -14.60%

2 YEAR   1.18%   -29.59%   -80.38%   -43.64%   -38.41%   -24.20%   -35.99%

3 YEAR   2.55%   -34.46%   -78.60%   -29.96%   -29.96%   -21.43%   -12.26%

4 YEAR   -4.50%   -21.45%   -62.27%   -26.90%   -10.25%   -25.64%   -2.04%

* Years based on weekly returns, last week of November 1998-2002.            

The CRP outperformed both the S&P 500 aswell as all the mutual fund portfolios exceptfor the four-year growth of FMSPX. More im-portantly, the S&P 500 outperformed most ofthe funds in all the time spans and where itdid not, the funds could not consistently beatthe S&P 500. While the CRP has outperformedthe S&P 500 for the past four years, over amore significant amount of time, they are sta-

tistically proven to be relatively equal. Theoutcome of this experiment suggests that evenin a relative short time span, passive invest-ing will outperform the active fund managers– the CRP’s “dartboard” approach and the S&P500 will beat the pros.

To further demonstrate Malkiel’s point thatstock returns are unpredictable, the CEO has

7

We can also use these results to examine theopportunity cost of attending Colby College.Consider the representative student trying tovalue their education at Colby after 4 years.Suppose she chose instead to take an amountequal to Colby’s comprehensive fee in 1999and ‘invested’ it in the CRP. Then in 2000 shedid the same, and again in 2001, and onceagain in 2002. This is not to suggest that acollege education is worth only what is spenton it, but instead, to merely illustrate how thelast 4 years tuition would have fended in themarket had it not been invested in education.

Had the amount of the four-year total compre-hensive fee at Colby been invested instead inthe CRP, her investment would have grown1.7%. Had the tuition been placed in the S&P500, her return on the investment would havebeen –23.92%.

run an ex-post forecast of the CRP and theS&P 500 over the last four years, using a ran-dom walk linear regression model. Accordingto the Random Walk Theory, the forecastshould be a basic linear projection, withoutvarying fluctuation. However, future datashould include a random disturbance term,meaning that actual future growth is essen-tially a random deviation from the forecast.

The actual growth for this span in the CRP was–4.5% while the forecast was for 34.9% growth– a staggering difference. The S&P 500 wasforecast to grow 116.8%; actual growth was –21.45%, as shown in the graph below. How-ever, these results are consistent with theRandom Walk Theory that historical informa-tion will not aid in the prediction of future stockperformance.

Random Walk As A Forecasting Tool

Using the same forecast method, growth inthe S&P 500 from November 2002 until No-vember 2006 is predicted to be 49.4%. Thisforecast seems excessively optimistic espe-cially given recent years’ sub-par performance.One of the factors leading to such a vast fore-casted boost in growth can be attributed tocertain stocks’ stellar forecasted performancesuch as ICE’s predicted 366.2% growth andGAP’s 130.3% increase. There is no reasonto assume that this forecast should be anymore accurate than the ex-post forecast. Thefuture of the equity market cannot be predictedwith any certainty, again giving the nod to thepassive investment strategy. Therefore, theCEO suggests buying young, accumulating di-versified equity indexes, as well as other in-vestment vehicles, and selling old. With a pas-sive investment strategy, investors are far moreprotected from market fluctuations, no matterwhat size, than with an active trading strategy.

8

    Colby    Annual Growth Rate   ReturnsYear   Fee   CRP   S&P 500   CRP   S&P 500

1999-00   31,580   -1.19%   18.53%   31,204   37,433

2000-01   32,750   -1.60%   -5.22%   62,928   66,516

2001-02   34,290   14.69%   -12.65%   111,501   88,059

2002-03   35,800   -7.20%   -17.44%   136,701   102,262

Total   134,420   -   -   136,701   102,262

* Colby Fee paid from previous years’ returns.            

* Years based on weekly returns, last week of November 1998-2002.    

* Compunded annually, based on yearly lumps sum fee payments.    

In Maine A Reason for Optimism

The Maine economy is currently at a cross-roads. After years of strong growth in grossstate product, state budget surpluses, andemployment, the economy has all but stalled.Many experts forecast this slowdown in Maineto worsen in the upcoming years. We see rea-sons for guarded optimism, however. Thereare fundamental economic and demographicissues facing Maine that are cause for con-cern, but we view Maine’s unique economicmakeup as a blessing that will ultimately res-cue the state.

Demographics

Population characteristics have a significantimpact on any economy. Maine’s population,in particular, sets it apart from the other statesin the US. Maine’s population is growingslower, older, unevenly, and non-diversely aboutthe state. For these reasons, Maine’s popula-tion proves especially troublesome for its eco-nomic well-being.

The population in Maine has grown slower thanthe U.S. as a whole for the past 130 years.Current projections indicate that populationgrowth in Maine will slow even further. Lowbirthrates and a high level of out-migrationamong young people are responsible for thisphenomenon.

Complementing the out-migration of youngpeople, aging “baby boomers” are also con-

tributing to Maine’s population problem. Thiscohort represents a significant percentage ofMaine’s population. In the 1990s, this groupwas of working age. Now, many of these Main-ers are approaching retirement. In order forMaine’s economy to continue to grow, ablebodies are needed in the workforce, and asthe “baby boomers” leave the workforce, theyput a strain on the economy. On the wholethis age group tends to save more, pouringless money into the economy. If this trendcontinues, higher health care costs and in-creased demand on housing and transporta-tion will have to be supported by the smallerpercentage of younger people.

Because of the appeal of a retirement in Maine,retirees of the baby boomer generation flockto the state, disturbing the age distributionbalance even further. Not only has the agebalance in Maine been disturbed, but the popu-lation density of the state has also been af-fected. Plentiful jobs attract people in thesouth while retirees flock to the coast. Virtu-ally all of the population growth has occurredin the south and along Maine’s coastal regions.This has put a strain on the state’s infrastruc-ture and threatens to choke off many naturalresource dependent rural communities in thenorth and east.

Another significant characteristic of Maine’spopulation is that it is the whitest state in thenation. Twenty-seven out of twenty-eight citi-zens are Caucasian. According to former Sec-

9

retary of Labor, Alexis Herman, “the lack ofMaine’s diversity will be a particular challengefor attracting business to the state.” Further-more, our own world becomes narrower whenwe lack the insights of people from differentcultures.

Total Employment in Maine

Maine’s population dynamics are a major rea-son for the recent lull in employment growth.However, in the coming years total employmentin Maine will begin to bounce back from thismost recent slowdown. During the 1990s, totalemployment grew significantly as the USeconomy was booming. Between 1992 and2000 almost 91,000 new jobs were createdin Maine. Beginning in 2001, employmentgrowth began to slow, increasing only by 0.93%but apparently Maine weathered the nationalrecession in decent shape. As we begin toemerge from this slowdown and the nationaleconomy gets back on its feet, total employ-ment in Maine will continue to grow but at sig-nificantly lower rates than during the late 90s.

Due to continued sluggishness in the nationaleconomy we saw relatively little employment

growth in the first three quarters of 2002, to-tal employment is forecasted to grow only by0.24% this year as a result. Employment isforecasted to grow at a slightly higher rate ofjust over 0.5% in 2003. We then foresee some-what higher growth toward the end of our fore-cast horizon, but dampened by the demo-graphic constraints mentioned above. In 2004,total employment is forecasted to grow by1.09% and in 2005 by 1.00%.

The unemployment rate in Maine has remainedconsistently below the national averagethroughout the national recession last year andall of 2002. By the end of 2002 the unem-ployment rate in Maine is forecast to be 4.2%,making for an average unemployment rate of4.0% for the year. In 2003, we predict that itwill rise slightly to 4.3%. As the post-reces-sion US economy adjusts itself, we will seethe yearly average unemployment rate drop to4.1% and 3.9% for 2004 and 2005, respec-tively. While manufacturing employment isforecasted to continue its steady decline ofthe past 2 decades, the non-manufacturingsector will pick up the slack. This redistribu-tion will lead to a lower overall unemploymentrate for the state of Maine.

500

520

540

560

580

600

620

640

1990 1992 1994 1996 1998 2000 2002 2004

Nonagricultural Employment in Maine

Thousands of Workers

Actual

Forecast

10

A Continuing Transformation for Maine’sEconomy

On top of the demographic issues, Maine con-tinues to experience a transition in the funda-mental structure of its economy. Two decadesago Maine still had a significant manufactur-ing sector. In 1950, one out of every two jobsin Maine was in the manufacturing sector.Today, that figure has dropped to one in nine –mirroring a nationwide decline in manufactur-ing. The closing of Hathaway shirt factory herein Waterville this year is just the most recentexample of the manufacturing sector’s moveout of Maine. Employment in this sector hasbeen declining for the past 20 years and isexpected to continue on this track as the ex-portation of manufacturing jobs continues. In2001, manufacturing employment in Mainedropped by 4.8%, the greatest amount since1991. This decline is forecasted to worsen in2002, with an anticipated decrease of 6.4%.In the coming years, the decline in manufac-turing employment is expected to continue, butat a somewhat slower pace. In 2003 and2004, manufacturing employment is fore-casted to fall by 1.36% and 0.61%, respec-tively. In 2005 the decline is forecasted toslow further, with employment in this sectorfalling by 0.15%.

Due to the movement of manufacturing indus-tries out of Maine, the economy has trans-formed into a service-sector based economy.The modest growth in total employment canbe attributed to an increase in non-manufac-turing employment, which grew rapidly duringthe late 1990s, but had slowed in 2001. In1998, 1999, and 2000, employment in thissector grew by 3.4%, 3.7% and 3.6%, respec-tively. However, in 2001, growth slowed to justunder 2%. Non-manufacturing employment isforecasted to increase by just 1.3% in 2002and growth will slip below 1% in 2003. Weforecast that this sector should recover some-what after that, with growth rates of 1.3% and1.2% in 2004 and 2005, respectively.

This transition from a manufacturing basedeconomy to a service-based economy has sig-nificant repercussions for Maine’s citizens. Themanufacturing sector’s contribution per job togross state product has steadily increased overthe past thirty years. The non-manufacturingsector’s contribution has remained constant.So, as more manufacturing jobs transform toother sectors, gross state product is negativelyaffected. Furthermore, the transition frommanufacturing to services often means lowerpaying jobs. Admittedly not all non-manufac-turing jobs are bad, and not all manufacturingjobs are good. Earlier this month in our dis-cussions with Governor King surrounding thisissue and our forecast for the Maine economy,he noted “…a job at MBNA [financial servicesector] is a lot better than a job plucking chick-ens [manufacturing/meat processing].” Clearlybeyond the economic ramifications, the changein the economy’s composition effects the livesof the people of Maine in many ways. None-theless, Maine’s new economy will continueto witness substantial changes in the compo-sition and trend of personal income in the yearsahead.

Personal Income

Given the changes in the composition of theMaine economy, we are projecting that the levelof personal income in Maine will increase by4.5% in 2002 followed by growth rates of 3.5%,3.7% and 3.6% for the years 2003, 2004, and2005, respectively. However, these annual in-creases are not as strong as they have beenover the past 2 years, which witnessed in-creases of 5.8% and 6.1% for the years 2000and 2001, respectively. The two key factorsthat determine the annual change in personalincome for Maine in our model are total em-ployment and personal income for the nationas a whole. With relatively little growth fore-seen in state employment growth, our fore-casts follow our assumptions for growth in na-tional income

11

20000

24000

28000

32000

36000

40000

44000

1990 1992 1994 1996 1998 2000 2002 2004

Personal Income in Maine

Millions of Dollars

Actual

Forecast

Retail Sales

Consumer retail sales have been a consider-able driving force for the Maine economy sup-porting the state’s relatively smooth transitionthrough the recent national recession, increas-ing 1.8 % in 2001 and 5.7% in 2002 (includ-ing a forecasted value for the last quarter of2002). Relatively strong retail sales have re-flected Maine’s employment and tourismgrowths of the past years. Zero-percent financ-ing and rebates on auto loans in the latter partof 2001 and in 2002 have also significantlycontributed to retail sales growth as Mainershave rushed to take advantage of the oppor-tunity to update their auto stock. Further ex-pansions in employment and tourism in com-ing years will strengthen retail sales, which areforecasted to grow by 2.4% in 2003, 2.2% in2004, and 1.8% in 2005.

Tourism

Although tourism is hard to measure, this in-dustry probably represents the largest em-ployer in Maine with some estimates placingthe total number of jobs at around 77,000.Also, it has been one of the fastest growing

industries of the state in the last decade. Lodg-ing sales reflect recent increases in tourismencouraged by aggressive state advertising,the natural beauty and varied landscape of thestate, as well as its extensive coastline. How-ever, as a result of the September 11th terror-ist attacks, the tourism industry sufferedacross the nation and consequently lodgingsales in Maine grew only by 0.2% in 2001.Tourism picked-up during 2002 with lodgingsales growing by 7.8% and is forecasted togrow by 3.1% in 2003, 1.5% in 2004 and 1.4%in 2005. The slight slowdown in the projectedrate of growth of lodging sales can be partiallyattributed to a forecasted depreciation of theCanadian dollar against U.S. dollar and thus adecrease in spending by Canadian tourists.

An key input assumption in our model concernsthe exchange value of the US dollar. We areforecasting that the US dollar will appreciateagainst the world’s major currencies, includ-ing the Canadian dollar, over our forecast hori-zon. Our model is based on a historical phe-nomenon known in academic circles as inter-est rate parity theory and forecasts a reactionof the exchange value of the dollar to changesin relative interest rates between the US and

12

the rest of the world. According to our forecasts,interest rates in US will increase as the economypicks up speed and the Fed begins to tightenmonetary policy. This will cause the exchangevalue of the US dollar to appreciate over the nextthree years, reaching an exchange rate of about1.7 Canadian dollars per US dollar by 2005.

Manufacturing Workers’ Weekly Hours

Manufacturing workers’ weekly hours will be af-fected as the manufacturing sector continuesto lose predominance in the Maine economy.Lay-offs from the decrease in manufacturingemployment will cause some inefficiencies in afirm’s production that will be offset through em-ployee overtime. Overtime is used because itcosts less to add hours to current employeesthan to replace laid-off workers. Overtime isreflected in the increase of average weekly hoursworked by manufacturing workers. Averageweekly hours are forecast to end 2002 at 42hours bringing the average for 2002 to 41.8hours. The following year, 2003 will keep thesame yearly average of 41.8 hours according toour projections. Slowly, the overtime for a 40-hour week is expected to decrease. Both 2004and 2005 have a lower forecasted yearly aver-age of 41.7 hours per week reflecting a slow

decline in the manufacturing sector and a risein worker efficiency as they adjust to fewer work-ers in the workplace.

Housing Permits

Maine’s rate of issuing housing permits has keptup with the national rate over the last few yearsand indicates a significant increase in 2001.With low mortgage interest rates and an economyon the rebound from a national economic slowdown, the start of 2002 featured a dramaticboom in the number of housing permits. Basedon three quarters of data, we expect to seegrowth of about 19% from 2001 to 2002.

In 2003, we should see a bit of a correctionfrom this large increase and are expecting a 5.5%decrease from the 2002 numbers. However,personal income is expected to rise steadilythrough 2005. The larger paychecks will enablemore Maine residents to own homes, and thusthe number of housing permits will rise. Thiswill be partially offset by increasing home mort-gage rates that are expected to reach 7.5% bythe end of 2005. Therefore, we expect the num-ber of permits issued to rise slowly after thecorrection with increases of 1% in 2004 and 3.5%in 2005.

13

*An

nual

rate

s of

gro

wth

, exc

ept a

s no

ted

** A

nnua

l ave

rage

s

14

The Colby Coincident Index of the MaineEconomy (CCI) is designed to provide a clearand up to date assessment of aggregate eco-nomic activity in Maine. The CCI is constructedas a weighted index of six economic indica-tors that reflect the changing structure of theMaine economy and current economic condi-tions in the state: Total Nonagricultural Em-ployment, Personal Income, Retail Sales, Av-erage Weekly Hours of Manufacturing Produc-tion Workers, Lodging Retail Sales, and TotalHousing Permits. A major advantage of theCCI is that it presents a picture of the Maineeconomy without the 2-year time lag associ-ated with Total Gross State Product (GSP). TheCCI allows for the study of recent trends in theMaine economy that cannot be as easily stud-ied using GSP. The CCI is generated using quar-terly data. GSP, on the other hand, uses an-nual data, and therefore cannot be used toinvestigate economic fluctuations that occurduring the year.

Our development of the CCI is based on themethodology used by The Conference Boardto compute their composite indices. The CCItakes into account the relative historic volatil-ity of each of the included economic indica-tors. Components with high variability areweighted less in order to capture trends inaggregate economic activity as well as smoothout expected variations in the particular com-ponents. As a weighted index of a variety ofindicators, the CCI offers a more complete pic-ture of economic events relative to single, morespecific, indicators that may lead to contra-dicting conclusions.

The individual components in the CCI werechosen based on their ability to explain pasteconomic activity while reflecting current trendsin the Maine economy. Lodging sales are notcommonly used in coincident indices, but wedecided to include in the CCI for Maine to cap-ture the economy’s reliance on tourism. The

lodging component also shows how the Maineeconomy is affected by economic fluctuationsoutside the state.

Because it is relatively stable compared to theother components of the CCI, non-agriculturalemployment is the most highly weighted com-ponent of the index. For the same reason, non-agricultural employment is also relied on asan indicator of changes in overall economicactivity. Housing permits, on the other hand,is given the lowest weight in the index due toits historical volatility. The CCI is indexed tothe first quarter of 1992 and is set equal to100 in that period.

Maine’s distance from the Silicon Valley techbubble that drove the recent recession, as wellas its low concentration of high-tech firms,served to insulate the state from the recentnational recession. According to the CCI, Mainedistinguished itself from the aggregateeconomy as it did not experience negativegrowth. However, this will also have conse-quences for growth in the future – as the USeconomy recovers from recession, growth inMaine will be lower than US GDP growth. TheCCI predicts that the Maine economy to growat a 2.6 percent pace this year, followed by aslower growth of about 1.5 percent annuallythrough 2005. We see the demographic prob-lems discussed earlier as the main obstaclefor rapid growth in Maine. An aging populationwhere youth and talent are drawn out of thestate is putting increasing pressure on theeconomy. Higher unemployment and lack ofan industrial base in the northern part of thestate will also restrict growth. On the positiveside, it bears noting that the tourism sector isshowing promise, which will help support fu-ture growth. Our outlook for the Maine economyis therefore moderately optimistic.

(For full CCI Forecast details see the next page)

The Colby Coincident Index of the Maine Economy

Predicting Presidents: Forecasting Maine’s Choice in 2004

An interesting economic forecasting applica-tion is to examine and predict other eventsthat are rooted in economic conditions. Tothis end Ray Fair, and economist at Yale Uni-versity, has created a model studying economiceffects on voting behavior in presidential elec-tions. The theme throughout this issue of theCEO has been the Maine economy. The fol-lowing special report is a digression from theforecast of economic sectors to predict howMaine will vote in the coming election. Byapplying our forecast of real GDP for the UnitedStates to a modified version of Fair’s model,we conclude that the Republican candidateshould have an easy time winning the state inthe 2004 election.

Fair looks at a number of economic and politi-cal variables to develop his model for predict-ing the percentage of the two-party popular votethat the Democratic candidate will receive. Theexplanatory variables of his model are whatparty is currently in power, whether or not theincumbent is running again, the annual percapita growth rate of the economy for the threequarters prior to the election, inflation over the

15 quarters prior to the election, the numberof periods where the annual growth rate of realGDP has been above 2.9%, the periods of timewhen the economy has been at war, and thenumber of terms a party has been in office.One lynchpin assumption of the model is thatall third party candidates draw evenly from bothcandidates vote totals. Our variation on Fair’smodel takes these same explanatory variablesand then regresses them with the two-partyvote in Maine since 1928.

For the 2000 election, the model predicted thatthe Democrat’s share of the vote would be54.4%, an overestimation relative to what ac-tually occurred of just 3.8%. This low errorprovides confidence that the model can accu-rately predict the two-party vote to within fivepercent. The forecast value for the 2004 presi-dential election is 0.290. In other words ourearly prediction for the next presidential elec-tion is that in Maine the Democratic candidatefor President will receive just twenty-nine per-cent of the two-party vote. If it’s true that “…asMaine goes so goes the nation,” then is a Bushlandslide in 2004 in the offering?

15

-2.1

3.0

3.9

1.61.8

1.41.5

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

4.0

1990 1992 1994 1996 1998 2000 2002 2004

Colby Coincident Index of the Maine Economy

Colby Econom

ic Outlook

c/o Associate Professor Michael D

onihueD

epartment of Econom

ics5

23

2 M

ayflower H

illC

olby College

Waterville, M

E 04

90

1