Questions about capital

Hello Jim, Q- I'm trying to build a dynamic IO model and am trying to figure out which information in the capital elements are applicable to that end. As such, I have some questions about the capital entry. What exactly is contained within capital? A- The capital column of the SAM shows investments by all industries (Capital payments to Commodity row) or borrowings by all institutions (Capital payments to Institution rows). Q-I know that it's the purchases of capital equipment and construction by industries and institutions, but what is included in this total? What relationship, if any, does capital have to BEA investment data? A- Capital Expenditures, or Gross Private Capital Formation (GPCF), are estimated at the national level using the BEA NIPA table 5.5.5 "Private Fixed Investment in Equipment and Software" bridged to IMPLAN sectoring through the BEA benchmark worktable that bridges the NIPA sectors to benchmark sectoring. The national GPCF is then distributed to states and counties based on total employment in all new construction industries. This implies that a purchase of capital goods within a state or county is linked to overall construction activity within that area. For example, a county containing 0.3% of the national construction employment would receive 0.3% of the national GPCF for all sectors except the construction sectors; for these sectors, national GPCF is distributed directly to the county or state based on TIO. Q-Does it only include non-residential purchases or are the purchases of homes (including by private households) included? A-It does include households. It does so in terms of Institutional borrowing and savings as well as home building. However, while household borrowing is in there, though it doesn't show how much of that borrowing is for a home purchase vs. other items. One indicator would be the levels of output in the New Residential Construction sectors in the area. Q-Are the purchases from non-capital producing industries (e.g. nondurable goods, retail sectors, etc.) because of the purchase of capital goods and construction from margin sectors? A- Yes - the purchases are margined, which is why you'll see "investment" in retail and wholesale sectors. Regarding non-durables, if the BM had an investment value in a given sector, we follow suit. These include some unexpected sectors, like services. However, the values are small compared to the more intuitive sectors. Also, a good thing to keep in mind is that R&D is now considered an investment rather than an intermediate expenditure, which is probably making up the bulk of those purchases in the service sectors (e.g., Custom computer programming services, Computer systems design services) in addition to the R&D sector itself. Does this address what you are trying to understand? If not, could you give us a little more detail about what you are trying to extract and translate into your dynamic Model? Regards, IMPLAN Staff

Thank you this does answer many of my questions. A few additional questions: 1) So industries invest and those investments are amalgamated and then disbursed to the producers of a given commodity whereas institutions borrow, correct? Are government and non-NAICS sectors included in the amalgamated industry totals? 2) As a follow up, how are public sector investments (including in equipment and software) in fixed capital treated in the SAM? 3) As far as the bridge table, are you referring to the 2002 Private Fixed Investment in Equipment and Software bridge table or is there a newer or different version that IMPLAN uses? Is this the source of the margins used? If so, are there any adjustments made to reflect the changes in relative prices or are the ratios constant from year-to-year? 4) Re: "The national GPCF is then distributed to states and counties based on total employment in all new construction industries." Is this from the BLS CEW/ES202 data? 5) Re: "Regarding non-durables, if the BM had an investment value in a given sector, we follow suit." What is BM referring to here? As far as the dynamic model, I'm just trying to construct a capital coefficients matrix (that is, the amounts of each industry's output held as a capital stock in the production of a given industry's output) using the data in IMPLAN and a few publicly available data sources (e.g. BEA's NIPA accounts, FRED data, etc.). See pp. 639-42 in Miller and Blair 2009.

Bump

Hello Jim 1) So industries invest and those investments are amalgamated and then disbursed to the producers of a given commodity whereas institutions borrow, correct? Are government and non-NAICS sectors included in the amalgamated industry totals? A- We recommend consulting the description of the SAM elements here: http://implan.com/index.php?option=com_content&view=article&id=294:294&catid=228:228. Capital investments by industry are represented in capital column payments to commodities in the SAM. The IMPLAN SAM doesn’t contain industry-level capital investment detail. It reflects the structure of the Bureau of Economic Analysis (BEA) benchmark input-output tables, in which capital investment is considered a final demand. Institutions both borrow and invest (save). Household payments to capital, for example, represent savings. Capital payments to households represent borrowing. The same logic holds for government payments to and from the capital account. Industry payments to Other Property Income, and the corresponding Other Property Income payments to the Enterprise row, reflects retained earnings that might be used for capital investment. Government capital investment is reported as its own columns in the SAM. 2) As a follow up, how are public sector investments (including in equipment and software) in fixed capital treated in the SAM? A- Our investment figures do indeed include public investment but not from that table. There are a variety of sources. Construction investment comes from NIPA Table 3.9.5. Government Consumption Expenditures and Gross Investment. SLGov non-structure investment comes from the Census Bureau's State & Local Government Finance data: http://www.census.gov/govs/local/ Control totals for total Fed and S&L Investment come from NIPA tables 5.1, 3.9.5, and 3.17, with help from the latest BEA Benchmark to break out into additional detail. 3) As far as the bridge table, are you referring to the 2002 Private Fixed Investment in Equipment and Software bridge table or is there a newer or different version that IMPLAN uses? Is this the source of the margins used? If so, are there any adjustments made to reflect the changes in relative prices or are the ratios constant from year-to-year? A-Out apologies it’s a reference to the PEQ bridge, which relates Nipa line codes to BEA input-output industry codes, and can be found here: http://bea.gov/industry/more.htm. The margin ratios are based on data from the 2007 benchmark. 4) Re: "The national GPCF is then distributed to states and counties based on total employment in all new construction industries." Is this from the BLS CEW/ES202 data? A- Non-construction investment is distributed according to construction output, which is distributed according to total value added and total employment, which includes proprietors. This article provides additional information about the data sets used to derive IMPLAN Employment: http://implan.com/index.php?option=com_content&view=article&id=450 5) Re: "Regarding non-durables, if the BM had an investment value in a given sector, we follow suit." What is BM referring to here? A- This is reference to BEA Benchmark input-output table As far as the dynamic model, I'm just trying to construct a capital coefficients matrix (that is, the amounts of each industry's output held as a capital stock in the production of a given industry's output) using the data in IMPLAN and a few publicly available data sources (e.g. BEA's NIPA accounts, FRED data, etc.). See pp. 639-42 in Miller and Blair 2009. " IMPLAN produces capital spending patterns, which show investment patterns by industry. This is based on a table on a “capital flows” table published by BEA (http://bea.gov/industry/capflow_data.htm). You can access the capital spending pattern by starting the Setup Activities window>Import>From Another Model>navigate to the “Utilities” folder>select one of the capital spending patterns, either with or without capital investment in structures. If you are interested in these, they can be downloaded from our website: 536 with structure: http://implan.com/index.php?view=document&alias=7-investment-with-structures-536sector&category_slug=536&layout=default&option=com_docman&Itemid=1764 536 FFE only: http://implan.com/index.php?view=document&alias=27-invenstment-only-ffe-536-sector&category_slug=536&layout=default&option=com_docman&Itemid=1764 Regards, IMPLAN Staff

Thank you for the detailed response. After I parse through all of the responses and sources, I'll follow up with any questions that arise. The only ones that stand out now is in response to your answer to question 5: a) Are these files available for the 509 scheme? b) Is there a version of the 440 that includes structures? c) Is there any way to extract these data directly using Access? Or am I better importing and then pulling it from the model impact events database? d) Are there any workarounds to the inability to import capital spending patterns in aggregated models? Thank you again!

Hello Jim, A - B: You can bridge 536 IMPLAN sectors to 509 and 536 to 440 using sheets available in our [url=http://www.implan.com/index.php?option=com_docman&view=list&slug=&Itemid=1764]downloads[/url] section. D: Consider importing the pattern, then aggregate the model rather than aggregating and then importing the pattern. C: Our user license does not support the access of our data in any software other than the IMPLAN system. Regards, IMPLAN Staff

My apologies, but I have additional questions: 1) In a previous response, you wrote, "Out apologies it’s a reference to the PEQ bridge, which relates Nipa line codes to BEA input-output industry codes, and can be found here: bea.gov/industry/more.htm. The margin ratios are based on data from the 2007 benchmark." I'm assuming this only holds true for model years which use the 2007 benchmark, is this correct? As a follow up, do you have a list of which years respond to which year's benchmarks (e.g. 2006 data corresponds to 2002 benchmark, etc.). I'm interested in model years 2000 - 2013. Also, you didn't answer as to whether the margin ratios are changed year-to-year to reflect relative changes in prices, are they? 2) Which table for private investment in structures is used in the capital columns commodity entries? My guess is that it's NIPA Table 5.4.5. Private Fixed Investment in Structures by Type, but I want to confirm this. Thank you for all of the help you've provided!

Hello Jim, The benchmark years are as follows: 2007 Benchmark = 2013-2014 data 2002 Benchmark = 2007-2012 data 1997 Benchmark = 2001-2006 data (no 2005 release) 1992 Benchmark = 2000 data We use NIPA table 5.3.5, Private Fixed Investment by Type, which we disaggregate with other data including Census' Value of Structures Put in Place (http://www.census.gov/construction/c30/historical_data.html). Based on this discussion, you may be interested in our new time series data. Our time series data stretches from 2001-2014 (including 2005) and utilizes the same metrics for every year along with updated data. The 2007 Benchmark is used for all data years and the sectoring scheme is consistent. If the time series data interests you, please contact us at 651 439-4421. Regards, IMPLAN Staff

I wish as it likely would have saved me countless hours. Unfortunately, I'm still a student and don't have the funding to purchase data sets. Thank you again for your assistance.

Does IMPLAN apply margins to the intellectual property products (IPP) categories in the Nonresidential fixed asset tables? Earlier you said that the PEQ bridge is used for the equipment sections and I was wondering if there's a similar bridge used for the IPP components. Most of them seem like they could be purchased directly (i.e. the services), but there are some manufacturing categories in there that might have margins. Some of them (e.g. Other computer and electronic product manufacturing, Motor vehicles, bodies and trailers, and parts manufacturing, etc.) overlap with PEQ categories. Does IMPLAN use the PEQ bridge for those items? Other items (e.g. Pharmaceutical and medicine manufacturing, Chemical manufacturing, excluding pharmaceutical and medicine, etc.) don't have PEQ analogues. Or are there no margins applied for the IPP categories?

Hello Jim, In general, the IMPLAN SAM is in producer prices. For commodities not covered in the PEQ bridge, we use the most recent BEA benchmark (2007 in this case), which is in producer prices (i.e., margined prices), to estimate capital investment. This includes most intellectual property products. Investment in motor-vehicle-related commodities and computer & electronic product manufacturing, as you point out, are included in the PEQ bridge. Accordingly, IMPLAN’s estimates are based on the PEQ bridge and Nipa investment data. Regards IMPLAN Staff