Benchmark I-O tables

Hi Ryan, Sorry about the delay in our response. Our understanding is that the BEA I-O tables are based off of Census business surveys. Most Census business surveys are mandatory, meaning that the Census will follow up with each establishment until they get a response. And yes, the Census does require establishments to report detailed economic figures. However, firm level data is confidential by law and never made publicly available. The industry categories that are reported are carefully designed to aggregate and obscure firm level data so that no piece of information can be traced back to an individual company. This is why there are instances where publicly released Census data is suppressed at the 5-6 digit NAICS level, or why there are issues with non-disclosures in some data sets. More information about Census business forms can be found here: http://www.census.gov/survey_participants/business_surveys/. Regarding the second part of the question, the technical coefficients for industry purchases are based on the national I-O model, while the technical coefficients for the local region are adjusted to local value added to output relationships and have imports removed. Local value added is based on state GDP data and county wage and salary and proprietor data. We use a gravity model to estimate regional purchase coefficients (RPCs) by commodity and then apply these RPCs to national industry purchases to regionalize the production function to the state or county area. More detailed information can be [url=http://implan.com/index.php?option=com_content&view=article&id=821%3Aresearching-implan-data&catid=185%3Adata-information&Itemid=166]found here[/url]. In summary, we know what regional Value Added is for every county and state in the U.S. and thus either shrink or expand the spending pattern coefficients to sum to the difference (i.e. based on the this equation Output = Intermediate Expenditures + Value Added, we can shrink of expand the sum of the coefficients of the Intermediate Expenditures (spending pattern) as needed. If national Intermediate Expenditures : Value Added ratio for a given industry is .75:.25 and reported Value Added for a state is .35, the Intermediate Expenditures are proportionately shrunk so that they sum to .65. If in another state, the Value Added component for that industry is .17 than the Intermediate Expenditures will be proportionately expanded to .83). Likewise we also regionalize the spending pattern by using data to estimate the amount of local demand met by local supply, thus allowing us to regionalize what can be purchased locally and what has to be imported to the region. Please let us know if you have any additional questions.