INTRODUCTION

IMPLAN is a widely used regional economic impact data and analysis application. Over the years, IMPLAN has gained the trust of academics and practitioners by providing valid and reliable data and analytical application. In this article, we’ll dive into the history of IMPLAN and of input-output analysis to show how we have come to be relied on as a credible source for researchers and decision makers.

INPUT-OUTPUT MODELS

IMPLAN relies on an economic modeling technique known as input-output (I-O) analysis. Wassily Leontief, who is also known as the father of I-O, won a Nobel prize in 1973 for his research on I-O analysis. A short description of Wassily Leontief can be found in this Investopedia article. A brief history of I-O is outlined in this article.

The foundational concept of I-O is that everything in an economy is interconnected through buy-sell relationships. The interconnectedness of industries is represented and measured through an Input-Output (I-O) Table. I-O models quantify the backward linkage effects of production, estimating economic impacts by analyzing which goods, services, and labor needs are required to produce products or services.

IMPLAN expands upon the traditional I-O approach to also include transactions between industries and institutions and between institutions themselves, thereby capturing all monetary market transactions in a given time period. IMPLAN can thus more accurately be described as a Social Account Matrix (SAM) model, though the terms I-O and SAM are often used interchangeably. I-O Tables and SAMs are static matrices that trace monetary flows throughout a region for a given period of time. The assumptions underlying I-O models are described in our support site article - Detailed Key Assumptions of IMPLAN and Input-Output Analysis.

IMPLAN

The IMPLAN system was originally developed within the U.S. Forest Service (USFS), and later expanded in partnership with the University of Minnesota, prior to becoming a privately-held company.  IMPLAN (IMpacts for PLANning) was the original name given to the system by the USFS. In fact, one of the USFS employees who helped develop the system later joined the company.

One of the two original founders continues to serve on IMPLAN’s Board of Directors to this day. Additionally, IMPLAN’s Chief Economist and Data Development Officer, Dr. Jenny Thorvaldson, has been with IMPLAN since 2010. She holds a PhD in Natural Resource Economics and has worked in the Data Department for the entirety of this timespan, thereby bringing stability and historical institutional knowledge to present-day IMPLAN. IMPLAN is proud to have two other PhD Economists: one in the Sales Department and another in Educational Services & Customer Success.

Today, IMPLAN is used not only by private businesses and non-profit organizations, but also by many government agencies and researchers at over 200 academic institutions. The flexibility of IMPLAN allows users to incorporate local-level expertise, and its functionality is applicable not only nationally but also internationally. As with any modeling system, the educated use of the model is the responsibility of the user. IMPLAN empowers its users with various training options and other educational resources and support.

From the tool’s inception to the current day, IMPLAN has dramatically expanded our offerings beyond the original Industry level data used for economic impact analysis. IMPLAN has added satellite accounts to the U.S. annual data, such as Environmental Data and Occupation Data, that allows users to bridge their economic impact Results to environmental impacts and occupations. Other additions to the IMPLAN data offerings include Expanded Demographics for U.S. regions in addition to Canadian Provincial and International Data Products.

The implementation of Data Library has also streamlined the process of exploring and comparing IMPLAN Data across regions and time. Data Library provides users with access to various key data assets at the county and state level throughout the United States for available IMPLAN Data Years (2001 - current year). This includes, among others, Industry Data, Employment & Wages by NAICS (Imputed CEW Data), and Tax Data.   

More details on our rich history can be found on the History page of our website.

IMPLAN DATA

Over time, the data development processes have undergone significant improvements as well, not only in terms of the knowledge and use of diverse data sets, but also in terms of the documentation of those processes, both internally and externally (articles and videos) and improved quality control procedures, including validation from the Education Services and Product Departments.

Because the data production processes can change somewhat from year to year, depending on the availability and quality of raw data sources and any possible  improvements that may be developed subsequent to quality control discoveries, IMPLAN re-estimates its entire set of models, from 2001 to the current data year, every five years, using current best practices and updated raw data sets,  yielding a truer time series.

With the exception of the domestic U.S. trade data, for which there is no raw data source, all raw data sources come from federal government agencies. In fact, the Bureau of Economic Analysis (BEA) acknowledges IMPLAN’s use of their national tables for regionalized models in their Input-Output Handbook, “Concepts and Methods of the U.S. Input-Output Accounts”.

IMPLAN’s domestic U.S. trade data show the trade of all goods and services between all sub-national regions, thereby allowing for a) more precise single-region impact estimates and b) multi-region impacts, whereby the indirect and induced impacts in other regions, associated with the direct impact to the base region by way of supply chains and commuting patterns, are estimated. These domestic trade flows are estimated via a double-constrained and calibrated gravity model, which is summarized in this video on our YouTube channel and described in full detail in this paper.

Additionally, the IMPLAN data constantly receive much scrutiny.  The Data Department spends a substantial amount of time investigating data values questioned by users.  Most often, such cases come down to a misunderstanding of the definitions (for example, total employment includes proprietors, not just wage and salary employees; household income includes all sources of income, not just wages and salaries; BLS CEW data do not fully cover all industries).  We at IMPLAN pride ourselves in not being a “black box” and are more than willing to help users understand and feel comfortable with our data and the functioning of the model.

VERIFYING RESULTS

The most critical step in any impact analysis is ensuring the direct inputs are modeled appropriately, as all related Indirect and Induced Effects stem from the initial change in the economy. The standard best practice for any impact analysis recommends an approach that would result in the most conservative estimate of positive economic effects, while also considering any negative impacts from the modeled change, otherwise known as a net effect analysis.

While one could never prove or disprove the estimates of indirect and induced effects from an impact analysis, attempts have been made to compare different modeling systems and to evaluate the relative accuracy of regional economic impact estimates. 

In general, economic models are designed to predict an outcome based on a set of variables constrained by assumptions about the relationships between them. The greatest limitation of economic models, and impact models in particular, are the underlying assumptions which are established to simply complex economic activity. While most economic impact models are subject to the same set of core assumptions, there are a few key differences from IMPLAN’s I-O model to a true CGE model, namely the use of forecasting in CGE models to predict future market conditions. 

Because other changes to the economy occur in addition to and simultaneous to the one being analyzed, an impact analysis cannot be performed ceteris paribus. While it is certainly reasonable to assume that certain aspects of a regional economy or industry may remain relatively constant from one year to the next, the lagged nature of IMPLAN data restrict a user's ability to define the current economic state of a region when estimating the overall impact of an economic event (Clouse, 2024). This is especially true when looking at the backward linkages, which change relatively little year over year.

Dudensing, Guerrero, & Amosson (2019) noted the importance of selecting the appropriate IMPLAN data year, or making customizations to the default IMPLAN regional models, to most closely mimic the economic conditions of when the modeled event(s) occurred in their ex post evaluation of a beef-pant closure. Remember, IMPLAN allows users to customize some of the underlying industry and commodity level data used by IMPLAN’s model to estimate effects, in addition to allowing users to fully customize the modeled event to match known activity. 

Lastly, when analyzing an event, the IMPLAN model assumes a period of one year, but an economic event’s actual impacts may take longer to be fully actualized. When attempting to validate results, consider using data collected over a period of time to perform a more accurate comparison of estimated versus actual results.

REFERENCES

U.S. Bureau of Economic Analysis (BEA). Concepts and Methods of the U.S. Input-Output Accounts. (Washington, DC: BEA, April 2009, 12-19) https://www.bea.gov/sites/default/files/methodologies/IOmanual_092906.pdf 

Bonn, M. A., & Harrington, J. (2008). A Comparison of Three Economic Impact Models for Applied Hospitality and Tourism Research. Tourism Economics, 14(4), 769-789. https://doi.org/10.5367/000000008786440148 

Clouse, C., Thorvaldson, J. and Jolley, G.J. (2023). Impact Factors: Methodological Standards for Applied Input-Output Analysis. Journal of Regional Analysis & Policy 53 (2): 1–14. https://jrap.scholasticahq.com/article/87960-impact-factors-methodological-standards-for-applied-input-output-analysis

Dudensing, R., Guerrero, B. and Amosson, S. (2019). Evaluating the Accuracy of Regional Economic Impact Estimates: Considering a 2013 Beef Plant Closure in Texas. Journal of Regional Analysis & Policy 49 (1): 92–107. https://jrap.scholasticahq.com/article/10204-evaluating-the-accuracy-of-regional-economic-impact-estimates-considering-a-2013-beef-plant-closure-in-texas

 

Written August 30, 2024