UW Department of Geography and the Washington State Input-Output Model

Submitted by Nell Gross on

Thank you very much to Professor Emeritus Bill Beyers for contributing this article! Update March 10, 2021: We are so happy to announce that Professor Beyers has recently received a commendation letter from Washington State Governor Jay Inslee for Professor Beyers's work on the state input-output table, applauding the "important contributions, ... technical skill and diligence and [Professor Beyers's] dedication as a public servant." Congratulations, Bill!

Input-Output models are portraits of national or regional economies that are widely used to calculate economic impacts, such as the impact of the Boeing Company on the Washington State economy, or the economic impact of arts and cultural organizations on the Central Puget Sound regional economy. They measure economic impacts through estimates of sales (output) by industry, labor income, and jobs created. They were developed in the 1930’s by Wassily Leontief, a Harvard University professor who won the Nobel Prize in Economics for his pioneering work on these models for national economies.

Washington State has a unique history of input-output models in the United States, and the University of Washington Department of Geography has been involved with in most of them. Nine models have been constructed, benchmarked against the years 1963, 1967, 1972, 1982, 1987, 1997, 2002, 2007, and 2012. These are all Economic Census years, a source providing baseline data for construction of these models. No other state has an history of locally-constructed models of this type.

Around 1960, geographers and economists started to construct these models for regions, and the Department of Geography was a pioneer in this work. Our strong economic geography specialization, led by faculty emeriti Edward Ullman and Morgan Thomas, moved the UW into the forefront on these models. Ullman published a monograph through the UW Press on the economic base of American cities, which he defined as the interregional traded-sector in input-output models. Thomas became part of the pioneering team that constructed the first Washington input-output model, benchmarked against the year 1963. I worked on this model as a research assistant for Thomas while a graduate student. Charles Tiebout in the Department of Economics was another pioneer on regional input-output models, and he helped with the construction of the 1963 model. Tiebout was also an adjunct professor in the geography department. I then chaired the team of faculty that developed the second model, benchmarked against the year 1967.

A great example of the use of input-output models for economic impact analysis stems from the recommendation of the North Cascades Study Team, appointed by President Kennedy, that a national park be established in the North Cascades. This recommendation was strongly opposed by the forest products industry, who hired faculty from the UW College of Forest Resources to use the input-output model to present estimates of economic impacts of reduced timber harvests due to national park establishment on the Washington State economy. The National Park Service asked Tiebout and myself to undertake a study of the economic impact of Mount Rainier- and Olympic National Parks to document the impacts of the spending of visitors to these existing parks. We started this study in 1968, but Tiebout died unexpectedly of a heart attack. We had to get approval from the federal Office of Management and Budget to conduct the study, requiring us to predict how many questionnaires would be returned by people surveyed to document their expenditures. We estimated a 25% return-rate, but we actually had over-65% return rate, testimony to the strength of public support for national parks. It took a long time to code all of these questionnaires, and while we completed the study, Senator Henry M. Jackson took the leadership in pushing legislation through the U.S. Congress to establish North Cascades National Park before our study was completed. The U.S. Forest Service and the forest products industry strongly opposed this legislation. Governor Daniel Evans went to the White House to urge President Ford to sign this legislation, which he did. Our study was completed in 1970, and it was widely cited in other battles over establishment of national parks.

The early Washington State input-output models used complex surveys to determine sales and purchases distributions of industries in the Washington economy. Professor Philip Bourque in the UW Foster School of Business was a leader in the development of these first models, and he recruited Richard S. Conway Jr. to help with the third model, benchmarked against the year 1972. Conway went on to get his Ph.D. at the University of Pennsylvania then returned to work for the State of Washington. He worked with Bourque on the 1982 Washington input-output model, but he was fired by Governor Dixy Lee Ray over her disagreement with an analysis he prepared using the input-output model. He then started his own consulting, firm, and became an affiliate professor in the UW Department of Geography, in large measure related to his work on the 1982 and 1987 input-output models. 

The State of Washington assumed responsibility for hosting the Washington Input-Output model on the Office of Financial Management (OFM) website beginning with the 1997 model. Conway developed the 1997 model based on the structure of the 1987 model. In 2002 and 2007 I again became the primary author of these models, working with a consortium of staff from Washington State government agencies. These models used extensive surveys to document sales and purchases distributions of Washington industries. Now I have been working with OFM to develop a new model for the year 2012 based on the structure of the 2007 model, which will be posted on the OFM website in 2020.