Generation and Interpretation of IMPLAN's Tax Impact Report

TAX IMPACT DETAILS

The table below clarifies the underlying level of detail of all line items in an IMPLAN tax impact report. In principle, the tax impact report captures all tax revenue in the study area across all levels of government that exist in that study area for the specific industries and institutions affected by an event or group of events. The underlying data that support the tax impact report, however, do not embody that much detail.

For example, IMPLAN does not have systematic reports of state government tax revenue by county; IMPLAN has same-year state government tax revenue by state and must allocate that to counties based on proxy information (we do have county-level data for some states, and use this to build a model for the allocation process).  Also, IMPLAN obtains detailed TOPI data by geography (even for each city within a county), but does not have any industry detail about the specific TOPI line item. A third note: for the data by city, we often must aggregate that to the county level, so that a model of two cities in the same county will have the same implied effective tax rates. In other words, city-specific data will be used, but averaged across all cities within a county.

Please note that all line items are controlled to nationwide, current-year controls estimated by the Bureau of Economic Analysis (BEA) in the National Income and Product Accounts (NIPAs) with no industry resolution and two level-of-government distinctions, Federal and State & Local. For example, the NIPAs might give a value of $15 billion in State & Local income tax in 2019, which would be reflected in the 2019 IMPLAN data.

Industrial and geographic resolution are reported at their maxima and nest more aggregate levels. For example, if IMPLAN has raw data on property tax at the county level, that implies we also have state-level data.

Timeliness lags are reported vis-à-vis the dataset year. For example, a 1 year lag for 2019 IMPLAN data means that the underlying data have a reference year of 2018. Timeliness is especially relevant for knowing whether changes in tax laws or economic conditions are reflected in the IMPLAN dataset.

The table below does not report the combined State & Local level of government, since (other than nationally in the NIPAs as explained above) IMPLAN does not collect any data at this level; it’s simply an aggregate for legacy and convenience purposes.

METHODOLOGY

Initially, the estimation of value added by sector and taxes by level of government proceed independently.

Taxes by level of government are obtained by combining data from the Annual Survey of State and Local Government Finances, which usually is lagged a year or two with respect to the IMPLAN data reference year, the most recent state government tax collections (also reported by the Census Bureau), and the most recent Census of Government Finance, which is like the Annual Survey, but covers every single unit of government. Those sources report tax by type, by unit of government (ergo by level of government), and by location. State government revenue is assigned only at the state level (i.e., the data do not tell us how much state income tax came from a given county). Federal government revenue is known only at the national level from the National Income and Product Accounts.

Data for county, city, and special district governments are assigned to the counties containing those units of government. Data for state and federal government revenue are allocated to counties based on proxies (e.g., personal income by county is used to allocate state government personal income tax revenue to counties). We have national level controls for taxes by level of government and type of tax from the National Income and Product Accounts. We first distribute taxes to states using a combination of the combined finances data and data on total taxes by state (covering both state and local governments) from the BEA’s Regional Economic Accounts. We then distribute those state values to counties based on the combined finances data, where possible, and by proxies where not possible.

The only industry-level detail we have for taxes is for TOPI, and it is for the entirety of TOPI, not the components of TOPI (e.g., sales tax, property tax, etc.). The components are only place- and level-of-government specific, not industry-specific. 

MISSING DATA

The raw Annual Survey of State and Local Government Finances data is missing data for some state- or county-level taxes in some areas. In this data set (and in IMPLAN's tax impact reports), the county-level column is distinct from the sub-county columns, in that it is not a sum of the sub-county columns but rather a distinct type of taxes.  So - a county can have sub-county types of taxes without having county-type taxes. The same is true for state-level taxes as that category is not a roll-up of county and sub-county taxes, but rather represents state-levied taxes distinct from county-levied, city-levied, and special district-levied taxes. So even though you know there are taxes levied in an area, they may not show up in the IMPLAN data. In this case, we recommend that you override the direct taxes in IMPLAN with your own estimates.

FILTERING

To filter by Direct, Indirect, and Induced taxes in IMPLAN, simply open the Filter window and click into the "Impact" Filter. This will provide you the option of "Direct", "Indirect", "Induced". Making a selection and clicking "Run" will apply the filter and only show the Tax Results as specified. if no selection is made, you are viewing total Tax Results. 

Updated December 13, 2021