Last night, we went to a Thai restaurant in Northampton, Massachusetts, the home of Smith College. Great food!

Today, Jon Lee from REMI continued yesterday's discussion of policy variables available in the REMI InSight model. He discussed the following blocks of variables: labor and capital demand; population and labor supply; wages, prices, and profits; market shares; and fiscal calibrations. Then, Adam Cooper, who is a REMI associate economist, described how to create alternative employment control forecasts.

In the afternoon, Stephen Houdlette from the University of Connecticut presented a paper, "Bradley International Airport Study Using TranSight Airport." This presentation reported research on the economic impact of Bradley International Airport, a major airport in Western Connecticut.

Fred Treyz, CEO of REMI, presented some information about REMI's new fiscal impact model, a dynamic economic and demographic model for assessing potential fiscal impacts of policies and events. For instance, what is the fiscal impact on the city of New Orleans of Hurricane Katrina. Will taxes rise. What will new operating costs be? To what level will county revenues rise due to rebuilding after Katrina. The fiscal impact model is wrapped around the REMI core model.

Last day tomorrow, and, then, back to State College!

From 28 through 30 September 2005, Rose Baker and I attended Regional Forecasting and Policy Analysis Workshop at the offices of Regional Economic Models, Inc. (REMI) in Amherst, Massachusetts. We attended the Workshop so that we could learn more about the general REMI regional economic model and specific new products offered by REMI for transportation and fiscal policy analysis.

We arrived in Amherst at about 2:00 am. Short night. The Workshop started at 9:00 am with a brief welcoming address by Fred Treyz, the CEO of REMI. Fred provided some context by discussing the recent hurricane Katrina disaster to frame some of the policy issues involved in economic analysis.As Fred said, "Unexpected events are to be expected." Fred emphasized that economic models are abstractions--complex abstractions, at that--of an even more complex economy. Economic models allow analysts to apprehend complex systems in economies.

Participants in the Workshop came from a diverse group of organizations. For instance, organizations represented included Northern Kentucky University, Illinois Department of Commerce and Economic Opportunity, North Country Council, University of Massachusetts, Team Northeast Ohio, Berkshire Regional Planning Commission, MIT, Connecticut Department of Commerce and Economic Opportunity, University of Connecticut, and, of course, Penn State.

Jon Lee, a REMI analyst, demonstrated the REMI Policy Insight economic model using a two-region example. Jon displayed and discussed the basic REMI variable elements for historical (1990-2002) and forecast periods (2003-2050). Jon used the two-region model to show the detailed baseline structure and outputs of REMI models using tabular and graphical information. Then, he worked through a simple simulation scenario in which he added employment to the tune of 500 workers in professional and technical services in the two-region economic area. Economic and demographic changes between the baseline and the simulation scenarios indicate the impacts of the added employment in the professional and technical services sector. Jon also demonstrated a REMI DevSight, a product that uses regional baseline information to develop a profile of the region(s) under analysis.

Another REMI analyst, Billy Leung, described the REMI product called TranSight, which is used in special applications that forecast highway/airport/rail financing, construction, and use. Travel demand data is included in Transight. Transportation projects involve construction, operation, and finance costs.VMT (vehicle miles traveled), VHT (vehicle hours traveled), and trip counts from some out-of-REMI-model estimation procedures (e.g., Tranplan, TransCAD) are input to TranSight.

Jon Lee started the afternoon with a presentation on analysis of multiple transportation corridors using REMI TranSight The project was a 22-mile road improvement in the state of Washington. This was a complicated analysis because much of the fiscal data seems so "squishy." Lots of imputation of the input data; but, small effects on the economy. I wonder how sensitive the findings are to error of imputation of inputs.

Two REMI folks, Adam Cooper and Julie Gressley (a Pennsylvania native!!), presented information about the analytical structure of the REMI Insight model. The absolute basics to understand the REMI model. Equations. Flow diagrams. Yahoooo! Love it. REMI Insight is, for my money, the most comprehensive analytical model of modern regional economies.REMI policy impact methodology requires, first, the construction of a baseline forecast--that is, a "status quo" forecast. Next, variables in the REMO model are changed on the basis of changes anticipated. For instance, employment in mining might increase if a new mine is opened in a region. Then, a policy forecast is created that includes the policy changes. The difference between the policy forecast and the "status quo" baseline forecast is the impact of the policy on the economy. The REMI model has many policy variables that the analyst can modify to portray the economic content of a policy initiative.

Day one of three. Looking forward to the remaining two days.

Over 22,000 visits to the Gulag Fun Park since its birth. Y'all are welcome at any time.

I Saw in Louisiana a Live-Oak Growing
--Walt Whitman

I saw in Louisiana a live-oak growing,
All alone stood it and the moss hung down from the branches,
Without any companion it stood there uttering joyous leaves of dark green,
And its look, rude, unbending, lusty, made me think of myself,
But I wondered how it could utter joyous leaves standing alone there
without its friend near, for I knew I could not.

My server at http://emerge.ed.psu.edu was hacked into. The IP address of the hacker resolves to a Romanian site. BUT, as we all know, hackers can travel through several sites before they land on yours. The server showed up in an IRC bot, with an indication that it was compromised and available. So, a trusted "white hat" hacker contacted Penn State with evidence that mine and 60+ other servers on the Penn State network were on this list of compromised servers. Our telecommunications people put a filter on the network that shut down my server and all others that were hacked. My server is not allowed online again until it is fixed. This fix will occur tonight and tomorrow (yawn!). No damage done to the world by the compromise, but lots of harm could result if left untreated. Ahhhh, cyberspace!