2010 to 2012 Changes in Induced Multipliers
I am updating an analysis that used 2010 IMPLAN data with 2012 data for the San Francisco Bay area region (9 counties) and noticed a significant decline in the Induced multipliers across the majority of industries. I have attached a spreadsheet to illustrate some examples. For instance the induced multiplier for sector 332 "Transp. by Air" decreased from .43 to .24, a 44% drop. This and other examples are contained on the tab "2010 vs. 2012." One thought I had was that employee household spending on web-based purchases out of state has increased which means that less HH spending is occurring locally. Do you have any other plausible explanations as to why there was such a large drop in the Induced multipliers? Is this related at all to changes in Proprietors Income? I included two additional tabs "Output Multipliers" that compares the 2010 and 2012 Induced multipliers for this region as well as "Industry Detail" for both 2010 and 2012 (same region). Thanks Adam
Hi Adam, Thank you for your post. One of the most important reasons multipliers tend to go down over time is increased productivity (i.e., increased output for the same amount of labor and non-labor inputs). There are many existing forum discussions about productivity: http://implan.com/index.php?option=com_kunena&view=topic&catid=84&id=17210&Itemid=1679#17222 http://implan.com/index.php?option=com_kunena&view=topic&catid=84&id=17621&Itemid=1679#17623 https://implan.com/index.php?option=com_kunena&view=topic&catid=80&id=18836&Itemid=1679#19080 Also, there was a second release of 2012 data. I recommend re-downloading your data and re-building your model/analysis. I believe the multipliers will look better (while most still do decrease a bit). Thanks!
Thanks for the Forum posts. I will review them for my questions. As for the new 2012 data, I re-downloaded the CA county files for 2012, re-created a new model for the 9 county Bay Area and the Output multipliers are exactly the same as the ones I previously sent you (see attached file, first tab). I assumed I would see some difference in the multipliers. Is there something I need to do to access the new 2012 data? Thanks Adam
There is no single or simple answer to why multipliers change over time, but here are some more considerations: Some more thoughts about a decline an industry’s induced multiplier: an industry might purchase less from local (i.e. within study area) sources and more from outside sources. If this happens, induced multipliers can change without there being any change in household behavior. Or, perhaps, industries make purchases from other industries whose labor income shares of output have declined. If proprietors that supply an industry in an area tended to earn less money in 2012 vs. 2010, this would be consistent with reductions in induced multipliers, since with new output, proprietors would be compensated according to the year’s average (which might be negative in some industries in some places – but more likely to be negative overall if the economy is slow). Regarding the relationship between proprietor income and induced effects: all else equal, lower or negative PI would result in lower induced effects multipliers, since proprietors are by definition within the study area and spend income via households (PI pays households, which then spend money, thereby creating induced effects). So, negative PI in any industry would be making the induced multiplier lower than it otherwise would be. LI per worker would change from increasing/decreasing average wages (not considering proprietors), increasing/decreasing average proprietor income, and changing employment counts (which in IMPLAN includes both wage & salary employment and proprietor employment). The effects of these income changes can be substantial. As we noted before, changes in profitability and productivity can significantly affect multipliers, too. Using Sector 46, Fats and oils refining and blending, as an example (you noted that it has a change in Type 1 Labor Income Type 1 multiplier of 3.36, i.e. it’s much bigger than it was): If you look at Study Area Data, you’ll see that EC and PI are about the same, but that Output was about half as much in 2010 as in 2012. Most of the difference in Output, when decomposed, goes to Other Property Income and Tax on Production and Imports. The Direct Effects Labor Income multiplier for 46 is .02 in 2012 and .05 in 2010. This is a direct result of much higher output/LI (and per worker, given that for this sector LI/worker didn't change much), i.e. productivity, in 2012 vs. 2010. That is the calculation that yields the Direct Effects Labor Income multiplier. Since the Direct Effects Labor Income multiplier is the denominator in the calculation of the Type 1 Labor Income Multiplier, one would expect a large increase if the denominator decreases at such a high rate. The difference between the Indirect Effects Labor Income multipliers between 2010 and 2012 is, proportionally, much smaller, so one can expect a large change in the Type 1 multiplier based on the study area data. Proprietor Income and OPI, and sometimes TOPI, all components of Value Added, can change significantly year-to-year. We have reviewed the multiplier changes in the spreadsheets you sent us, and we find them to be within plausible ranges of changes we could expect over 2 years. If you look at the individual direct, indirect, and induced effects, the changes appear much less drastic, even in the industries that appear to have large changes in Type 1 multipliers. Here are a couple old forum posts on multiplier changes, as well: http://implan.com/index.php?option=com_kunena&view=topic&catid=80&id=17736&Itemid=1679#17805 http://implan.com/index.php?option=com_kunena&view=topic&catid=80&id=17403&Itemid=1679#17437 Please let us know if you have more questions on the topic.
Please sign in to leave a comment.