2013 vs. 2012 Data Agricultural Data
We have been conducting analyses on the contribution of agriculture for a few Midwestern states. We had purchased 2012 IMPLAN data in December 2013 in anticipation of the 2012 Census of Agriculture being released in early 2014. To provide more timely analysis for our clients we used the 2012 Census of Ag to calibrate the IMPLAN data. We did this knowing that the IMPLAN data did not yet reflect the 2012 Census of Ag. Briefly, we adjusted output values for primarily production agriculture (commodities such as oilseeds and grains) to better reflect the actual amounts. We then applied the original VA coefficients to the new sales amounts and calculated new VA category amounts. We also ensured the absorption coefficients were adjusted as appropriate.
So my question is, now that the 2013 data are out and at least the processes I've described above have been complete, what do you recommend for data purchases of other Midwestern states? We do want to have some comparability to what we've done in states we've already completed the work for, but realize that there may be some time savings realized by purchasing the 2013 data. Any help is appreciated. Thanks.
-
Hi Spencer, While the decision on how to proceed is up to you, here are some considerations. The 2012 dataset as shipped from IMPLAN reflected 2012 state-level output amounts for agriculture based on the USDA’s Economic Research Service published data. Their method is current and reasonably detailed at the state level, but is not as comprehensive as the Census of Agriculture. So, for the year 2012, the Census of Agriculture would have the best estimates for total output. For the 2013 datasets, we used the 2012 Census of Agriculture to do a couple of things: 1) allocate state-level output of crops to counties and 2) disclose state-level values not published by ERS (or the USDA National Agricultural Statistics Service - NASS), but we did not use its state-level output estimates because we had more up-to-date options. Additionally, we improved our 2013 state-level output data by integrating two data sources (instead of just 1) that have empirical measurements of state-level agricultural output: NASS and ERS. Both of these sources publish 2013 output values. So, the 2013 IMPLAN data will have the most up-to-date county-level agricultural output values available and will have 2013 output values (versus 2012 values from the Census of Agriculture). The 2013 IMPLAN data output values for crops won’t be consistent with 2012 Census of Agriculture insofar as the 2013 IMPLAN data will have newer output values (2013 instead of 2012). If comparability is paramount, you do have the option of redoing old analyses with 2013 IMPLAN data. This would be a trade off of a cost of some extra time, but would save time of doing more manual adjustments. The 2013 data would offer several benefits of having newer county distributions, more up-to-date economic relationships, and 2013 estimates of state output. Otherwise, sticking to 2012 data and the method for incorporating the 2012 Census would be best. Note that if the original 2012 IMPLAN datasets were purchased in December 2013, they are likely Release 1 (of 2), and so moving to 2013 data would be even more beneficial. You could also use the 2013 IMPLAN data sets (which will have up-to-date distributions among counties, among other enhancements), and revise study area output for crops back to 2012 values if wished, but the rest of the model will generally reflect the economy in 2013. Thank you! -
Hi Spencer, While the decision on how to proceed is up to you, here are some considerations. The 2012 dataset as shipped from IMPLAN reflected 2012 state-level output amounts for agriculture based on the USDA’s Economic Research Service published data. Their method is current and reasonably detailed at the state level, but is not as comprehensive as the Census of Agriculture. So, for the year 2012, the Census of Agriculture would have the best estimates for total output. For the 2013 datasets, we used the 2012 Census of Agriculture to do a couple of things: 1) allocate state-level output of crops to counties and 2) disclose state-level values not published by ERS (or the USDA National Agricultural Statistics Service - NASS), but we did not use its state-level output estimates because we had more up-to-date options. Additionally, we improved our 2013 state-level output data by integrating two data sources (instead of just 1) that have empirical measurements of state-level agricultural output: NASS and ERS. Both of these sources publish 2013 output values. So, the 2013 IMPLAN data will have the most up-to-date county-level agricultural output values available and will have 2013 output values (versus 2012 values from the Census of Agriculture). The 2013 IMPLAN data output values for crops won’t be consistent with 2012 Census of Agriculture insofar as the 2013 IMPLAN data will have newer output values (2013 instead of 2012). If comparability is paramount, you do have the option of redoing old analyses with 2013 IMPLAN data. This would be a trade off of a cost of some extra time, but would save time of doing more manual adjustments. The 2013 data would offer several benefits of having newer county distributions, more up-to-date economic relationships, and 2013 estimates of state output. Otherwise, sticking to 2012 data and the method for incorporating the 2012 Census would be best. Note that if the original 2012 IMPLAN datasets were purchased in December 2013, they are likely Release 1 (of 2), and so moving to 2013 data would be even more beneficial. You could also use the 2013 IMPLAN data sets (which will have up-to-date distributions among counties, among other enhancements), and revise study area output for crops back to 2012 values if wished, but the rest of the model will generally reflect the economy in 2013. Thank you! -
Hi Spencer, Here is an article that further discusses the second release of the 2012 data: http://implan.com/index.php?option=com_content&view=article&layout=edit&id=906 Thanks!
Please sign in to leave a comment.
Comments
4 comments