Volume Author/Editor: Charles L. Ballard, Don Fullerton, John B. Shoven, and John Whalley. Volume Publisher: University of Chicago Press

Transcription

1 This PDF is a selectin frm an ut-f-print vlume frm the Natinal Bureau f Ecnmic Research Vlume Title: A General Equilibrium Mdel fr Tax Plicy Evaluatin Vlume Authr/Editr: Charles L. Ballard, Dn Fullertn, Jhn B. Shven, and Jhn Whalley Vlume Publisher: University f Chicag Press Vlume ISBN: Vlume URL: Publicatin Date: 1985 Chapter Title: Data n Intermediate Prductin and Value Added Chapter Authr: Charles L. Ballard, Dn Fullertn, Jhn B. Shven, Jhn Whalley Chapter URL: Chapter pages in bk: (p )

2 Data n Intermediate Prductin and Value Added 4.1 Intrductin In bth this and the fllwing chapter we describe ur 1973 benchmark data set. We describe in sme detail the raw data and the many adjustments in rder t assist anyne wh wishes t recnstruct r mdify any part f the data set. In this chapter we deal with data n the use f factrs by industry, taxes by industry, input-utput transactins, and gvernment enterprises. Chapter 5 is cncerned with data n husehld incmes and expenditures, general gvernment expenditures, and freign trade. Value added cnsists f net payments t the factrs f prductin plus factr taxes. We use the Natinal Incme and Prduct Accunts (NIPA) t btain figures fr capital incme, labr incme, and tax by industry in We listed the nineteen industries used fr ur prductin data in table 3.1. These nineteen industries are a direct aggregatin f the fifty-nine industries given in detailed tables f the Survey f Current Business (SCB) r f the eighty industries given in Bureau f Ecnmic Analysis (BEA) input-utput tables. These, in turn, are based n the Standard Industrial Classificatin (SIC). Appendix A f this chapter shws hw we aggregate the eighty BE A industries. 1 This disaggregatin prvides us with richer detail than many f the tables published in the SCB, which nly reprt infrmatin fr twelve industries. Fr this reasn much f ur data cmes frm unpublished wrksheets f the Cmmerce Department's Natinal Incme Divisin (NID). 2 Even with this greater degree f disaggregatin f the prductin sectr f the ecnmy, we are 1. The relatinship between ur classificatin and the SIC can be seen in a table in U.S. Department f Cmmerce, BEA 1975, p The staff at NID were extremely helpful in prviding us with detailed wrksheets. We als received excellent suggestins fr additinal surces and alternative adjustments prcedures. 55

3 56 Chapter Fur left with fairly aggregated grups. We have tried t maintain separate classificatins fr industries that are quantitatively imprtant and taxed differently. While chapter 6 describes verall adjustments t the data that are necessary t meet general equilibrium cnsistency requirements, this chapter describes many adjustments that were made t the raw data fr definitinal cnsistency. Hwever, tw discrepancies cannt be crrected here. First f all, the Natinal Incme Divisin cllects sme data n an establishment basis and ther data n a cmpany basis. Fr ur purpses, the best measure is incme generated in the prductin f each gd, but data are nt available n this basis. 3 The secnd prblem is that the accunts use a "natinal" definitin f ecnmic activity, where we wuld prefer a "dmestic" definitin Labr Incme and Labr Taxes by Industry The grss-f-factr-tax return t labr is the sum f emplyee cmpensatin and the estimated return t the labr f self-emplyed individuals. The NIP A definitin f emplyee cmpensatin includes wages and salaries and supplements t wages and salaries. 5 These supplements include emplyer cntributins fr scial insurance (ECSI) and ther labr incme (OLI). Emplyer cntributins include thse fr Old Age, Survivrs, Disability, and Hspital Insurance (OASDHI r Scial Security), unemplyment insurance, public wrkmen's cmpensatin, and railrad retirement. Other labr incme includes private wrkmen's cmpensatin, pensins, grup health and life insurance, and supplemental unemplyment benefits. The return t the labr f self-emplyed persns is an unknwn fractin f the ttal return t the entrepreneur wh invests his time and capital jintly. NID prvides us with data fr the ttal return t unincrprated enterprises by industry. Hwever, we must still estimate the prtin f this return that accrues t labr. In rder t estimate this return, we assume that partners r prprietrs in an industry earn the 3. "The establishment basis is used fr the industrial classificatins f wages and salaries, supplements t wages and salaries, incme f unincrprated enterprises and inventry valuatin adjustment, and interest paid by nncrprate enterprises. Hwever, the cmpany basis is used fr crprate prfits, the crprate inventry valuatin adjustment, and interest paid and received by crpratins" U.S. Department f Cmmerce, OBE 1954, p The natinal definitin measures "the incme and prduct attributable t factrs f prductin supplied by residents f the cuntry rather than the incme and prduct f factrs physically lcated in the cuntry" ibid., p Wages and salaries include all mnetary remuneratin f emplyees, including the cmpensatin f crprate fficers, cmmissins, tips, bnuses, and receipts in kind that represent incme t the recipients. These are n an accrual, nt disbursement basis. Definitins f this kind can be fund in U.S. Department f Cmmerce, BEA 1976a, pp

4 57 Intermediate Prductin and Value Added same return as ther wrkers in that industry. We derive the average cmpensatin f emplyees in each industry frm tw tables in the July 1976 SCB (U.S. Department f Cmmerce, BE A 1976b). Next we divide emplyee cmpensatin by the full-time equivalent number f emplyees (prviding average annual labr incme fr each industry). Then, using NID data n the numbers f prprietrs and partners, with a crrectin fr hurs wrked, we estimate the imputed return t the labr f self-emplyed individuals. 6 Fr several industries cnstructin, utilities, real estate, and manufacturing the estimated labr cmpnent exceeds the ttal unincrprated incme. This prblem is less serius in manufacturing, because unincrprated enterprises nly prduce a small prprtin f manufactured gds. Hwever, large segments f cnstructin, utilities, and real estate are unincrprated. These three industries are land intensive and prbably have a high rate f accrued capital gains. These gains wuld nt shw up in the figures fr unincrprated incme. Rather than assign a large return t labr and a negative return t capital in these industries, we simply assign the ttal unincrprated return t labr. The sum f the imputed return t labr in the unincrprated sectr, and wages and salaries in the crprate sectr, yields grss-f-tax labr incme fr ur nineteen industries. These appear in clumn (1) f table 4.1. We define taxes n labr incme t include emplyer cntributins fr scial insurance (ECSI), emplyee cntributins fr OASDHI, plus cntributins fr OASDHI by self-emplyed individuals. We include all taxes n labr incme that discriminate amng industries. OASDHI has a fixed rate up t a maximum per emplyee, s that high-incme emplyees pay a smaller tax per dllar f labr incme than emplyees whse incmes fall belw the maximum. Thus industries with higherthan-average cmpensatin f emplyees tend t pay a lwer effective rate n OASDHI. Unemplyment insurance als discriminates by industry, since the tax is greater fr thse firms with a higher incidence f unemplyment. We treat cntributins fr public wrkmen's cmpensatin (part f ECSI) as a tax, althugh we treat cntributins fr private wrkmen's cmpensatin (part f OLI) as rdinary incme. This distinctin is smewhat arbitrary, since bth are mandatry insurance payments. We make the distinctin because the cntributins fr wrkmen's cmpensatin g 6. If this imputed return t labr is subtracted frm ttal unincrprated incme, the residual is an estimate f the return t capital. Alternatively, using an asset basis, if we can measure the capital stck f bth the crprate and nncrprate sectrs in each industry, we can attribute t nncrprate capital the same net-f-tax rate f return as that f crprate capital. Subtracting this estimate f return t capital frm the ttal unincrprated incme yields a residual return t labr estimate. In practice, the latter (asset basis) methd is mre difficult, with capital stck by industry and by sectr generally unavailable.

5 58 Chapter Fur t gvernment. NID supplies us with breakdwns f ECSI items by industry. Since the Scial Security prgram invlves matching cntributins, we use the emplyer share t estimate the emplyee share. 7 The ttal fr cntributins by self-emplyed individuals is given in the July 1976 SCB. Hwever, NID des nt divide this ttal by industry. We allcate the ttal amng industries by the prprtin f self-emplyed labr incme in each industry. The ttal tax n labr fr each industry is shwn in clumn (2) f table 4.1. We subtract this frm grss-f-tax labr incme t get labr incme net-f-tax, which is clumn (3). Using clumns (2) and (3) f table 4.1, we calculate the effective tax rates n labr by industry, which are shwn in clumn (4). The rates differ by industry, but they are clse t the verall weighted average f 10.1 percent. This dispersin is nt great fr tw reasns. First, ur prcedure fr estimating self-emplyed cntributins t OASDHI by industry actually assumes the same rate f tax in all industries. Hwever, these cntributins represent nly $2.5 billin ut f $63 billin f labr taxes. Secnd, when we cmbine detailed industries int less-detailed nes, we tend t reduce the dispersin f tax rates. The rate f.0702 fr agriculture, fr example, reflects the fact that many agricultural wrkers are nt cvered by insurance prgrams. The rate f.0903 n services is due t a lw level f cverage and sme high salaries abve the ceiling fr OASDHI. 4.3 Capital Incme and Capital Taxes by Industry The return t capital, net-f-factr taxes, includes crprate prfits after tax, the return t unincrprated capital, net interest paid, and net rents paid. We wuld like t include real accrued capital gains in ur measure f capital incme in each industry, but the data are nt available. (Mrever, cnstancy f relative prices is cnsistent with ur parameterizatin f an equilibrium ecnmy with n excess prfits.) 8 Nw we will discuss the varius cmpnents f capital incme Crprate Prfits after Tax The July 1976 SCB gives crprate prfits after deducting prperty taxes and indirect business taxes. SCB table 6.21 shws crprate prfits after the crprate incme tax. This crrespnds t a net-f-tax definitin. This series includes prfits after the capital cnsumptin allwance 7. This step is justified empirically, since 1973 emplyer cntributins t OASDHI ttal $30,549 millin, while emplyee cntributins ttal $30,388 millin. (See U.S. Department f Cmmerce, BEA 1976b, table 3.11.) 8. Since we include crprate prfits after tax, we accunt fr any capital gains n shares that might result frm real retained earnings.

6 59 Intermediate Prductin and Value Added Table 4.1 Labr Incme, Tax, and Effective Rates by Industry in the U.S. fr 1973 Labr Incme Grss f Tax (1) Tax n Labr (2) Labr Incme Net f Tax (3) Effective Tax Rate (2) - (3) (4) All industries Agriculture, frestry, fisheries Mining Crude petrleum and gas Cntract cnstructin Fd and tbacc Textiles, apparel, and leather Paper and printing Petrleum refining Chemicals, rubber, and plastics Lumber, furniture, stne, clay, glass Metals, machinery, instruments, and miscellaneus manufacturing Transprt equipment and rdnance Mtr vehicles Transprtatin, cmmunicatins, utilities Trade Finance and insurance Real estate Services Gvernment enterprises 708,037 17,398 5,182 3,723 56,216 18,823 19,715 20,944 3,073 21,344 19,349 97,163 15,131 16,422 65, ,984 36,000 8, ,964 16,723 64,997 1, ,308 1,859 2,268 1, ,957 1,930 9,167 1,393 1,358 6,188 13,745 3, ,179 1, ,040 16,257 4,718 3,415 50,908 16,964 17,447 18,996 2,834 19,387 17, ,738 15,064 59, ,239 32,839 7, ,785 15, Nte: All figures are in millins f dllars. Cmpnent detail is available upn request. (tax depreciatin) but befre the capital cnsumptin adjustment (crrectin t ecnmic depreciatin) r the inventry valuatin adustment (IVA). The capital cnsumptin adjustment will be treated separately later. We btained the IVA frm NID f Cmmerce. 9 Clumn (1) f table 4.2 shws prfits after crrectin fr the IVA. 10 The ttal is $45,633 millin The negative $301 millin IVA fr mining and crude petrleum is divided between the tw by their prprtins f value added in The negative $178 millin fr "ther" crprate industries is allcated entirely t services, since this is the largest cmpnent, with a majrity f inventries. 10. The relatinship between these NIPA crprate prfits and the IRS definitin can be seen in U.S. Department f Cmmerce, BEA 1976b, table 8.5. The NIPA starts with IRS numbers and adds depletin, bad-debt adjustment, and ther dividends, and the cst f issuing securities, in rder t reach a measure related t current prductin. 11. Federal Reserve Bard (FRB) earnings are included in crprate incme figures, and FRB payments t the Treasury are cunted as part f the crprate incme tax. This

7 en v en t t en r^ c*^ io H (N O M N 00 H O f- en ON en Tf > i 0 t 00 en ON ON n N r i * ON r, ON 00 m en in (N en v en ON <N O <* rj- * <N v t-» en ON in ON 00 ^H CN en T-H in r» ON v en s I I I 2 '5 b g *«U cu H enenv mvenenin CN en en en ON IT) 1 in 00 r- (N ON ON 00 en en r t t- VO I u I I a es U restry, fisheriei tries re, f a 3 2f C O G Alii Agri Mini trleu d gas act cnsti ucti and tba< appar d leather i printing l refirling ical:s, rubber, E furnit ure, stne, cla; lachin ery, instru scellaileus mam prt : equi]pment and rd r velhides prt atin, cmmuni, utilit c CO E t *^J c Cruc Cm r \ c CO <U 773 cd Text Pape G 3 l lastis a. a P Chei E +2 c 1 Met; 3 CO 'E CO c Trar Mt( Trar ca u C CO 'SJ _ OJ 3 nd im v> CO C Trad Fina «J Real Serv eises (U UIUJ (U O

8 61 Intermediate Prductin and Value Added Capital Cnsumptin Adjustment Neither published nr unpublished surces at the Cmmerce Department give sufficiently disaggregated industry data n the capital cnsumptin adjustment r n "ecnmic" depreciatin. In the aggregate, the Cmmerce Department uses straight-line depreciatin fr cnsistent accunting and a replacement cst basis t measure ecnmic depreciatin. Then they subtract frm the capital cnsumptin allwance (as measured by the Internal Revenue Service) t get the capital cnsumptin adjustment. We tk capital cnsumptin adjustment figures fr manufacturing industries frm Cen (1980) and aggregated them t ur industrial classificatin. These figures are shwn in clumn (2) f table 4.2, alng with nnmanufacturing data taken frm the Cmmerce Department Return t Unincrprated Capital The prcedure used t calculate these numbers has been described in the earlier sectin n labr incme. We impute a return t the labr f prprietrs and partners, and subtract this frm the ttal t yield a residual capital return. If the imputed labr share is t large, we assign capital a share f zer. The results are shwn in clumn (3) f table Net Rents Paid Rents paid by an industry are payments fr "brrwed" capital, prperty, buildings, and machinery. The Internal Revenue Service (IRS) and the NIPA treat these payments as a cst deductin fr the renting firm and as incme fr the wners f the rented prperty. Since we seek t measure all payments t capital used in each industry, we treat rents paid as capital incme in the paying industry. Fr each industry we wuld like t add t capital incme all payments fr rental prperty and subtract rental earnings. This wuld be equivalent t adding net rents paid. The NIPA has n industry distributin fr rents paid, and the IRS Statistics f Incme (SOI) shw nly grss rents paid by industry. Hwever, the NID des prvide the ttal f net rental incme, which is equal t the ttal net rents paid. This ttal and its cmpnents are shwn in table 4.3. We treat net rents frm farm realty as a return t capital used in industry 1 (agriculture, frestry, and fisheries). The NIPA measure f imputed net rent frm wner-ccupied and ther dwellings sum t $12,917 millin. This figure is added t the capital incme f ur real estate industry. peratin is nt cvered by the crprate incme tax system and shuld be excluded frm calculatins n that tax. FRB earnings d make up part f capital incme in ur finance and insurance industry, hwever, and their payments t the Treasury can best be mdeled as part f a tax n capital incme in that industry, but nt part f the crprate tax.

9 62 Chapter Fur Table 4.3 Net Rental Incme Cmpnents, 1973 Surce Amunt 1. Net rents frm farm realty 3, Imputed net rent frm wner-ccupied dwellings 10, Net rents frm ther dwellings" 2, Ryalty earnings 1, Net rents frm business and gvernment b 3,124 5a. Nnfarm business 2,884 5b. Nnprfit rganizatins 59 5c. U.S. gvernment 160 5d. Freign gvernments Ttal net rental incme frm the private sectr 0 21,237 Nte: All figures are in millins f dllars. Items 1 thrugh 5 were btained frm NID and include the capital cnsumptin adjustment fr an ecnmic measure f depreciatin. "The sum f items 2 and 3 is net rents frm husing, equal t $12,917 millin, fund in table 1.20 f the July 1976 SCB. These include farm husing. b Iterns 5a thrugh 5d are an apprximate breakdwn f item 5. See text fr prcedure. The sum f items 1 thrugh 4, 5a and 5b. Similarly, cpyrights, patents, and ryalties paid fr natural resurces shuld be cunted as capital incme in the industry where these capital assets are used and paid fr. This item 4 in table 4.3 ttals $1,498 millin, but we are nt supplied with industry divisin. The SOI shw nly ryalties received. We therefre resrt t a prcedure used by Rsenberg (1969, p. 152), and allcate 12.5 percent t printing and publishing, 12.5 percent t electrical machinery (phngraph recrds), and 75 percent t natural resurces. We apprximate the use f natural resurces by industry frm the depletin deductins taken fr tax purpses in 1973 (U.S. Department f the Treasury, IRS 1977a and 1977b). We assign deductins prprtinally by depletin t six industries: petrleum and natural gas, ther mining, petrleum refining, lumber, primary metals, and utilities. 12 Finally, NID supplies us with net rents frm business and gvernment. These ttal $3,124 millin, as shwn in table 4.3. Net rents paid by each grup are unavailable, s we use the prprtin f grss rent paid t divide the ttal net rent paid. 13 Nnprfit rganizatins are part f the services industry. We attribute the estimated net rents f these institutins directly t the services industry. The $2,884 millin estimated net rental payments f business were 12. Rsenberg (1969) distributes 75 percent amng three il- and gas-using industries, despite the existence f "ryalty" payments fr lumber and ther mining. 13. Grss nnresidential rent paid by business and gvernment was $45,235 millin. The prtins paid by nnfarm business, nnprfit rganizatins, the U.S. gvernment, and freign gvernments were als prvided by NID. We used these data t calculate items 5a t 5d in table 4.3.

10 63 Intermediate Prductin and Value Added allcated amng industries by the prprtin f grss rent paid by industry, furnished in the crprate and business SOI. This prcedure generally fllws Rsenberg (1969), and the results are shwn in clumn (4) f table 4.2. The NIPA imply that all f these rents riginate in the real estate industry. The NIPA real estate industry includes husing services, ther real prperty, and intermediatin services. We redistribute the secnd cmpnent, but are still left with a smewhat curius definitin f the real estate industry: husing services and intermediatin in real prperty Net Interest Paid The amunt f an industry's capital return that is paid t bndhlders will be reflected in that industry's net interest payments. We have used detailed wrksheets by industry, prvided by the NID. The wrksheets shw interest flws fr The dllar payments f interest by each industry are referred t as net mnetary interest paid. Since all kinds f prducers issue bnds t raise capital, these payments are psitive fr all industries except ne. Finance and insurance (F&I) has a large negative value fr net mnetary interest paid, due t the peculiar structure f that industry. The F&I industry engages in financial intermediatin, thus its interest receipts exceed the interest it pays n depsits. These net interest receipts are the frm in which the F&I industry receives payment fr services rendered. We argue abve that net interest paid shuld be added t capital earned in each industry. If we add the (negative) net interest paid t the prfits f the F&I industry, ttal capital is measured by a negative number. In 1973, net mnetary interest paid by F&I was negative $61,181 millin. A cnsistent slutin t this prblem exists if we assume that F&I cmpanies d nt issue bnds t raise funds. That is, we assume that n interest payments are made fr the capital used in the prvisin f intermediatin services. 14 In fact, we can raise the F&I net-interest-paid figure t zer by imputing additinal interest payments t ther industry and t persns. Thus, we have tw separate kinds f interest payments: mnetary payments and imputed payments. This view f the wrld 14. Savings and lan assciatins in 1973 paid $12 billin in interest n depsits and $1 billin n brrwed mney. Mst f the latter went t the Federal Hme Lan Bank Bard, which means that the savings and lan assciatins paid fr brrwed reserves at certain times. It is nt bvius hw t classify this kind f brrwing: it is brrwed capital used t prvide financial services, but it is nt fr building plant r equipment as a factr f prductin. Cmmercial banks paid $20 billin interest n depsits, $4 billin n federal funds (brrwed reserves), $.5 billin n ther brrwing, and $.25 billin interest n capital ntes and bnds. Only the last categry is clearly analgus t interest paid by ther industries, but it is such a small prtin f F&I net interest paid that it can be safely ignred. These tw types f institutins had the nly substantial payments n brrwed mney.

11 64 Chapter Fur attributes all prperty incme f F&I straight thrugh t the depsitrs wh als make (imputed) service charge payments t F&I. 15 The Cmmerce Department NID has made sme f these imputatins, but they nly ttal $41,702 millin. This falls $19,479 millin shrt f the ttal fr net mnetary interest paid by F&I. There are several reasns fr the difference. Fr instance NID des nt make an imputatin fr finance cmpanies and small-business investment cmpanies. We have chsen t frce the net interest paid by F&I (including bth mnetary and imputed interest) t be zer. T d this we impute an additinal $19,479 millin f interest paid by F&I. These interest payment imputatins shuld be allcated t ther industries accrding t their use f financial services. As a prxy fr the use f financial services, NID distributes the imputed interest receipts in prprtin t cash held (including demand depsits) by each industry. These data cme frm the SOI. We use the same prprtins t allcate all $61 millin f imputed interest, including the $19 billin that are nt imputed by NID. Based n these prprtins, we allcate 79 percent f the ttal t persns and gvernment. Out f the amunt that we allcate t persns and gvernment, sme $15,388 millin (r 79 percent f the $19,479 f extra imputed interest) is nt cunted by NID. The result is a GNP increase f $15,388 millin. N ne is better ff, but ur accunting shws larger cnsumer interest incme, ffset by larger (imputed) service charge payments t the finance and insurance industry. Net interest paid, bth mnetary and imputed, is shwn in clumn (5) f table Special Treatment f Depletin Deductins T btain prfits fr tax purpses, natural resurcefirmsin 1973 culd deduct a percentage f revenues t reflect the depletin f their reserves. The cmbinatin f these depletin deductins and expensing f explratin csts is a case f duble cunting, because in cmpetitive equilibrium, explratin csts are matched by the value f expected discveries. We can avid this duble cunting by adding depletin deductins t IRS incme figures. Data in the Statistics f Incme indicate that these deductins ttaled $9,301 millin fr crpratins and $429 millin fr unincrprated enterprises in The Cmmerce Department adds the nncrprate figure t its incme statistics, but adds nly $5,828 millin f the crprate figure. The difference ($3,473 millin, which equals 37 percent f crprate depletin deductins) is allcated as fllws: We take crprate deductins fr mining and lumber times.37 and add the result t incmes in thse industries. We take the ttal f crude petrleum and petrleum-refining deductins times.37 and add this t 15. This treatment is cnsistent with cmpetitin and the absence f abnrmal prfits. If an F&I cmpany is superir at discvering investments fr its clients, it can make larger service charges.

12 65 Intermediate Prductin and Value Added incme in crude petrleum alne. We d this because the incme statistics fr petrleum-refining crpratins are n a cmpany basis, s their deductins are fr crude petrleum peratins. These adjustments are reflected in clumn (1) f table 4.2* Tax n Capital Incme In ur mdel, the tax n capital incme at the industry level has three cmpnents the crpratin incme tax levied by the federal gvernment, crpratin franchise taxes levied by the state gvernments, and prperty taxes levied at the state and lcal levels. We use capital incme as the tax base fr all three taxes, even thugh the legal tax bases fr the latter tw taxes are capital stck and capital assets, respectively. Infrmatin n the crpratin incme tax cmes frm table 6.20 f the July 1976 SCB. We list this infrmatin in clumn (1) f table 4.4. Infrmatin n crprate franchise taxes, frm an unpublished wrksheet frm NID, is displayed in clumn (2) f table 4.4. There are n gd natinal statistics fr the prperty tax by industry. The best data fr 1973 are cntained in a wrksheet frm NID. The NID data are cllected frm a wide variety f surces. Fr sme industries data are gathered n yearly prperty tax payments frm publicatins like Agriculture Statistics, Life Insurance Fact Bk, and Statistics f Cmmn Carriers. Prperty taxes in the manufacturing industries are estimated frm a prperty tax survey in the 1958 Census f Manufacturing. In sme cases NID tk the estimated prperty tax rate fr ne f these industries and applied it.t an SOI estimate f the assets f anther industry fr which n direct infrmatin n prperty taxes was available. In the remaining industries NID built up an estimate f prperty taxes fr the entire natin by taking a weighted average f the prperty tax rates f the states. We must make ne adjustment t the NID prperty tax figures. In adjusting ur figures n net rents paid, we mve sme incme frm the real estate industry t ther industries where industrial real prperties were used. The prperty tax figures frm NID still include tax n thse prperties in the real estate industry, s we redistribute a part f the real estate industry's tax t ther industries. Rather than assume the real estate industry tax rate n the incme reassigned, we assume the rate f the industry where the prperty is used. First, we calculate a subttal fr capital incme withut net rents paid. Then, we calculate effective prperty tax rates in each industry and apply these rates t the net rent reassigned frm NID's real estate industry. The prperty taxfigureswith this adjustment are given in clumn (3) f table 4.4. Clumn (4) f table 4.4 shws the sum f the three capital taxes. Ttal net-f-tax capital incme is shwn in clumn (5) f table 4.4. The effective tax rates (clumn 4 divided by clumn 5) are shwn in clumn (6),

13 «x " -Iz e 13 5 S'H. O rt rt H E- U T-I r T-I (N.a, g- x < H r (N ON CO OJ,s«^8 ON in r~ T3 C II 3 «^ "^ 3 OO «. ca H r= H "75 c -a "3 O X <u _ * <" O O a T3 S P 60 i* C C <U Z V -g X) 3 s * - -a ".a «all*.. «3 J3 3 PL, 0. U J a> c3 )«a "3 3 C T3 s a fl - 3 S s s S S 8 «^ S IB H * ^ a > <U 03 f3 O - " 1/3 C ^ c3 H E a *j O

14 67 Intermediate Prductin and Value Added where.5270 is the average rate fr all industries. Agriculture has the lwest effective tax rate and transprtatin equipment has the highest. 16 We use capital incme net f all taxes t reflect the use f capital in each industry, but we shuld nte that measuring the industrial use f capital services by capital incme is nt the nly methd available. Several studies, including thse by Kendrick (1976) and Jrgensn and Sullivan (1981) have cllected data n capital stck by industry. In a perfect risk-free equilibrium with perfect measurement, bth data sets wuld shw the same distributin amng industries. Capital wuld be allcated s that the rati f net capital incme t value f capital stck is the same in every industry. Withut a perfect equilibrium measurement, hwever, we culd still ignre the capital incme figures derived here, use capital stcks, and attribute the same rate f return t all industries. Such alternatives have been investigated, but they d nt make a majr difference t the results f simulating alternative tax structures in the general equilibrium mdel. 4.4 Persnal Factr Taxes In sectin 3.3 we described the cnstruct we use t mdel the features f the persnal incme tax (PIT) which discriminate by industry. Equatin (3.31) shws hw persnal taxes are cllected at an average persnal marginal rate T n a fractin/ f the i th industry's payments fr capital (CAPi). This weighted average persnal marginal tax rate is 27.8 percent. The persnal factr tax (PFT), is thus equal t CAP^T. Crrectins at the persnal level ensure that individual grups pay their wn marginal rate, T ;, (/ = 1,..., 12) n an average fractin,/, f capital incme frm industry. T calculate the/, we use data n the varius types f capital incme by industry. These include crprate prfits (dividends and retained earnings), net interest payments (mnetary and imputed), net rent payments (including the imputed net rent frm wner-ccupied hmes), and the return t capital used in nncrprate business. These types f capital incme are treated differently by the persnal incme tax. In ur mdel we calculate the prprtin, g, f each type f capital incme that is fully taxable by the persnal incme tax. An industry's / is the weighted average f these g prprtins. The weights are the amunts f these capital incme types in each industry. The frmula fr / is given as DIV igd + RE igre + INT igl + MRENT igmr + IRENT igm (4l) _ + NCI ignc - (NCITQ/T) where the variables are defined in table 4.5 and discussed further belw. 16. The next sectin describes hw we add the persnal taxatin f capital incme by industry t the effective tax rates f table 4.4.

15 68 Chapter Fur The last term in the numeratr f equatin (4.1) is TrK igcg /y. If TT is the rate f inflatin with n relative price changes, then nminal appreciatin in the i th industry is equal t IT times the capital assets used there. 17 We use 7 percent inflatin fr 1973, frm the Ecnmic Reprt f the President (Cuncil f Ecnmic Advisers 1973). In ur mdel, K t are measured in terms f capital service units, each f which earns a dllar per perid in the benchmark. A real after-tax rate f return, 7 =.04, is used t cnvert between capital stcks and capital service units, as described in sectin 3.4. Then Kj/y is measured in capital assets, and TrKig CG /"y is the taxable prtin f nminal capital gains. The denminatr f (4.1) is equal t CAP h defined in sectin 3.3 as capital incme net f crprate and prperty taxes. Our capital usage by industry, K {, is net f persnal factr taxes: (4.2) K i = CAP i (l-f i T). Thus the final term in the numeratr, which invlves K h implicitly includes a term in^. T slve fr f t by itself, define A,- - DIV igd + RE igre + INT igl + MRENT igmr (4 ' 3) + IRENT igm + NCI lgnc - (NCITQ/T) ; then equatin (4.1) can be rewritten as K } Ji Slving fr^, ne btains CAP, (4-5) ^ Table 4.5 summarizes the parameter values we have chsen t implement these equatins. In the case f dividends, the federal gvernment lst an estimated $285 millin f revenue due t the $100 dividend exclusin frm the PIT in 1976 (U.S. Cngress 1977). We divide this by T t get an estimate f nntaxable dividends, equal t $1,024 millin. Since ttal dividends paid were $24,631 millin in that year, the prprtin taxable was This figure is used fr gd. We assume that retained earnings are reflected in appreciated values f 17. Relative price changes in 1973 might have prvided real capital gains as a majr cmpnent f capital incme in land-intensive industries. Hwever, the return t land is largely included in ur return t capital already. Our capital incme includes imputed returns t wner-ccupied husing, and actual rents paid in all industries, and the prfits f crpratins that wn the land they use. Only real appreciatin f idle and vacant land wuld be excluded frm ur capital incme figures.

16 69 Intermediate Prductin and Value Added crprate stck. Thus retained earnings are subject t sme persnal taxatin thrugh the capital gains tax. We use g RE t dente the effective fractin f retained earnings that is fully taxed by the PIT. Martin Bailey (1969) has prvided evidence that an average dllar f retained earnings leads t ne dllar f capital gains, in present value terms ver the lng Table 4.5 Variables in Equatin (4.1) Used t Define/, the Prprtins f Capital Incme Taxable at the Persnal Level Symbl Definitin Value r Surce DIV ( Dividends paid by the i th industry g D REi g RE INTj g f MRENT t g MR IRENTj g IR NCIj g NC NCITCj T TT Prtin f dividends taxable by persnal incme tax (PIT) Retained earnings f the i th industry Prtin f retained earnings taxable by PIT Net mnetary and imputed interest paid by the i th industry Prtin f interest taxable by PIT Net mnetary rent paid by i th industry Prtin f mnetary rent taxable by PIT Imputed rent earned in the i th industry Prtin f imputed rent taxable by PIT Nncrprate capital incme in the / th industry Prtin f nncrprate capital incme taxable by PIT Nncrprate investment tax credit in the / th industry Capital weighted average f persnal marginal tax rates Inflatin rate Ki Use f capital service units in the / th industry 7 Cnversin frm capital service units t capital assets units g CG Prtin f nminal capital gains taxable by PIT Frm table 6.22 f the July 1976 SCB.9584, as discussed in text Denned as crprate prfits (crrected by the IVA, depletin, and capital cnsumptin adjustments f sectin 4.3) minus DIV t.25, as discussed in text Clumn (5) f table 4.2, as discussed in sectin , since interest is fully taxable Clumn (4) f table 4.2 (except real estate), as discussed in sectin , since net rents are fully taxable Zer fr all but real estate, equal t $10,334 millin as shwn in table , since n PIT n wner-ccupied husing Clumn (3) f table 4.2, as discussed in sectin , since NCI is fully taxable Described in text and shwn in table , derivatin in equatin (3.31).07, fr 1973 frm the Ecnmic Reprt f the President Frm 1973 capital incme net f all tax.04, frm the real after-tax rate f return as described in the text Differs by industry, as described in text

17 70 Chapter Fur run. 18 There are, hwever, tax rate and deferral advantages t this frm f capital incme. In 1973, gains deferred until death were nt taxed at all. Bailey has shwn that clse t ne-half f lng-term capital gains is realized in a relatively shrt perid, while the remainder is held fr varying duratins averaging perhaps thirty-five years r mre. Weighing the advantages f deferral and these bservatins regarding hlding perids leads us t the cnclusin that a tax n $.25 f regular incme wuld yield apprximately the same persnal incme tax revenue as a tax n ne dllar f retained earnings. Therefre, the g RE fr retained earnings is set at.25. A value f 1.0 is assigned t the g parameters fr interest, mnetary rents, and nncrprate incme, since these types f capital incme are fully taxable by the persnal incme tax. Imputed net rents, which appear nly in real estate, get a g m f zer because these are untaxed. Since the nncrprate investment tax credit reduces the persnal incme tax (PIT) liability, we include in the numeratr f equatin (4.1) the amunt f incme that wuld result in the reduced tax liability. It is the amunt NCI (NCITC/T) that, when multiplied by T, yields tax cllectins f NCI T NCITC. The statutry rate f investment tax credit fr equipment was 7 percent in Estimates f the nncrprate investment tax credit (NCITC) here are intended t apprximate the amunts that wuld have accrued by sectr in 1973 if a 10 percent credit had been in effect and if there had been n limitatins n the use f the credits. The Treasury Department's Office f Tax Analysis prvided us with these estimates, and their prcedures can be summarized as fllws. The ttal 1973 NCITC estimate f $1,078 millin is scaled up by 10/7. T allcate this ttal amng the industries, it is assumed that the rati f ITC t depreciatin is the same fr the crprate and nncrprate prtins f each industry. Ttal nncrprate depreciatin is taken frm the IRS nncrprate Statistics f Incme. Each industry's crprate depreciatin is shwn in the July 1976 SCB. Using the Treasury Department's crprate ITC estimates fr each sectr, the rati f ITC t depreciatin is applied t the ttal nncrprate depreciatin. This gives an estimate f each industry's NCITC. This vectr is scaled up slightly t the prper ttal f $1,540 millin, shwn in table 4.6, clumn (1). The final parameter t discuss is g CG. Because nminal capital gains receive the same deferral advantages as real capital gains, we set g CG t.25 fr all industries except husing and agriculture. Fr husing we nte that, in value terms, 73 percent f residential structures are wner ccupied. 19 We assume that wner-ccupants effectively avid the capital 18. Fr an alternative theretical view, see Auerbach 1979b, Bradfrd 1980, and King Unpublished natinal balance sheet frm the Federal Reserve Bard.

18 71 Intermediate Prductin and Value Added Table Nncrprate Investment Tax Credit (in millins f 1973 dllars) and fi Parameters fr Each Industry Industry Nncrprate ITC fi Parameters (1) (2) Agriculture, frestry, fisheries Mining Crude petrleum and gas Cntract cnstructin Fd and tbacc Textiles, apparel, and leather Paper and printing Petrleum refining Chemicals, rubber, and plastics Lumber, furniture, stne, clay, glass Metals, machinery, instruments, and miscellaneus manufacturing Transprt equipment and rdnance Mtr vehicles Transprtatin, cmmunicatins, utilities Trade Finance and insurance Real estate Services Gvernment enterprises All industries , gains tax n their husing, whereas landlrds pay taxes n 25 percent f their gains. The rate f inclusin n all f husing is thus nly (.27)(.25) = In the agricultural sectr mst farms are privately held and are seldm exchanged in a taxable manner. We have set g CG in the agricultural sectr t.10. The resulting values fr all benchmark fi are shwn in clumn (2) f table 4.6. Nte that the fi parameter is less than ne fr mst industries. This reflects the lw persnal taxatin f retained earnings and the nncrprate investment tax credit. The fi fr real estate is.376, reflecting the nntaxatin f imputed net rents frm wner-ccupied hmes. Smefi are greater than ne, hwever, because f the taxatin f purely nminal capital gains. The use f capital by gvernment enterprises (industry 19) is assigned an fi f.75. This reflects rughly the prtin f interest payments that is taxable by the persnal incme tax. Apprximately ne-furth f gvernment's interest payments are n nntaxable state and lcal bnds We assign an fi f 1.0 t the use f private capital by general gvernment. This simplificatin aids in ur cmputatins, but des nt affect general gvernment's capital tax rate, equal t the average industry tax rate (see sectin 3.5).

19 72 Chapter Fur Finally, in table 4.7, we use the fi parameters t derive the amunt f persnal factr tax in each industry, shwn in clumn (3). When all capital taxes are subtracted frm capital incme, we have net incme f clumn (4). The last clumn shws the effective tax rates that we use in the mdel the ratis f all taxes t net capital incme in each industry. Since the tax rate is defined as taxes relative t net incme rather than t the mre cmmn use f grss incme, the rates can exceed unity. 4.5 Input-Output Transactins Data Our mdel requires data n interindustry transactins, that is, each industry's purchases frm each ther industry. We cnstruct an interindustry transactins table frm tw separate tables f The Detailed Input- Output Structure f the U.S. Ecnmy, 1972, published by the Bureau f Ecnmic Analysis (U.S. Department f Cmmerce, BEA 1979). The first table is the "make" table, which gives the dllar values f the prductin f each cmmdity by each industry. The secnd is the "use" table, which gives the dllar values f the use f cmmdities by industries. Our gal is t transfrm these tables int a single table shwing each industry's use f the utputs f the varius industries. T illustrate the prcedure, let us cnsider the case with nineteen industries and eighty-five cmmdities. The make matrix, M, is thus 19 x 85 and the use matrix, U, is 85 x 19. The first step in arriving at a transactins table is t define a 19 x 85 matrix, M, which is equal t M with the clumn sums nrmalized t unity. The (i, j) element f M is the prprtin f the ttal prductin f cmmdity / that is prduced by industry i. The transactins table, T, (19 x 19), is given as (4.6) T u = 1 M ik U k,fr. " " " k=\ j i.,..., iy Figure 4.1 gives an example f the prcedure fr a hypthetical case f tw industries and three cmmdities. The resulting 1972 interindustry transactins table reprts these transactins in millins f dllars at prducer prices. The valuatin at prducer prices excludes distributin csts. The trade margin and transprtatin csts appear as inputs t each industry frm the trade and transprtatin industries. It is imprtant t nte that nly the trade margin is included as an input frm the trade industry. The input-utput tables d nt trace the actual flws f cmmdities t and frm the trade sectr. If trade were shwn as buying and reselling cmmdities, ther industries wuld be shwn as buying mst f their inputs frm the trade industry. Instead, cmmdities are shwn as mving directly t the users, bypassing the trade sectr.

20 & H W H s O X ca ca O f N O O O O O O O SI ca ca fa H X f f ( N a >u-)0 NNN>n>in O ON * t U O 2 * ON ON 00 S ca vc * 1^ 2»S 60 (5JQ _r a c ca.s q3 u c c ca 0) ca 1> x at ca D h ft. Oi ^ y.ca G 5 JS OH si 2 -e 5 «c «) d ca ill I = ' 3 M CO u u -2 "S a ~ > «a Z - «-4-» C W *-* c- 3 n) ^ w Q rt fl g H H E(2^6 c * 6 I II

21 74 Chapter Fur IdAKE MATRIX Cmmdity A B C Industry 1 $10 $30 $30 2 $15 $10 $20 1 NORMALIZED ] V1AKE MATRIX Cmmdity A B C USE MATRIX 1 2 INTERINDUSTRY TRANSACTION MATRIX 1 2 Industry x A B C $20 $20 $10 $10 $20 $ $29 $28 $21 $17 Figure 4.1 Example f cnstructin f interindustry transactins table frm make table and use table. Table 4.8 United States Interindustry Transactins, 1972 (in millins f dllars at prducers' prices)

22 75 Intermediate Prductin and Value Added The aggregated 19 x 19 matrix fr 1972 appears in table The elements, (i, /'), f this matrix are the amunts f the utput f industry i that are used by industry/. Thus, if we lk dwn the fifth clumn, we see that the fd industry buys mre frm agriculture, frestry, and fisheries ($37,437 millin) than frm any ther industry. Clumns (8), (13), and (17) als prvide intuitive stries: petrleum refining uses much crude petrleum, the mtr vehicles industry is a heavy user f primary and fabricated metal and machinery, and real estate purchases a large amunt frm the cnstructin industry. At this level f aggregatin the elements alng the diagnal f the matrix are quite large. If we were t use a mre disaggregated industry classificatin, these diagnal elements wuld be relatively smaller. Finally, this (19 x 19) transactins table is updated t Each clumn is multiplied by the rati f 1973 value added t 1972 value added fr that industry. These ratis are based n unpublished data frm the Cmmerce Department's Natinal Incme Divisin. Since value added data are nt available fr the gvernment enterprises industry, we use the weighted average rati fr the eighteen private industries. Table 4.9 cntains all nineteen ratis. The nly rati far frm the average is 21. The crrespndence between the classificatin f the riginal matrix and ur nineteen industries is in Appendix A f this chapter , , , ,

23 76 Chapter Fur Table 4.9 The Rati f 1973 t 1972 Value Added in Each Industry Industry Rati 1. Agriculture, frestry, fisheries Mining Crude petrleum and gas Cntract cnstructin Fd and tbacc Textiles, apparel, and leather Paper and printing Petrleum refining Chemicals, rubber, and plastics Lumber, furniture, stne, clay, glass Metals, machinery, instruments, and miscellaneus manufacturing Transprt equipment and rdnance Mtr vehicles Transprtatin, cmmunicatins, utilities Trade Finance and insurance Real estate Services Gvernment enterprises agriculture, frestry, and fisheries. The large increase in value added in this industry is due largely t the sharp increase in grain prices that ccurred at that time. 4.6 The Matrix f Transitin between Prducer Gds and Cnsumer Gds The nineteen gds prduced by ur industry divisins d nt crrespnd t the cmmdities purchased by cnsumers. Fr example, cnsumers make very few direct purchases f the utputs f the mining and crude petrleum industries. The same is true f the primary metals and machinery industries. Trade services and cmmdity transprtatin services are nly purchased indirectly by cnsumers. The aggregate cnsumptin vectr fr final demands f husehlds, derived frm table 2.6 f the July 1976 SCB, cntains fifteen cmmdities rather than the nineteen industry utputs. The ttal is adjusted t exclude expenditures in the United States by freigners. We als adjust the ttal t include the $15,388 millin imputed service charge payments frm cnsumers t the finance and insurance industry, which was discussed in sectin 4.3 abve. After these adjustments, the ttal f $827,525 millin in cnsumptin expenditures n fifteen cnsumer gds crrespnds directly t the $827,525 millin cnsumptin demand fr nineteen prducer gds. The

24 77 Intermediate Prductin and Value Added latter vectr represents the industry utputs that are used t frm the fifteen cnsumer gds. T accmmdate these different classificatins, we treat these prducer gds as being cnverted int cnsumer gds by a fixed-cefficient Z matrix, estimated frm table B in the February 1974 SCB. This table used 1967 input-utput data and can easily be transcribed t an 86 x 84 matrix where the (/, rri) element is the ttal amunt f prducer gds i used in cnsumer gd m. The eighty-six industries used by the Bureau f Ecnmic Analysis aggregate directly t ur definitin f nineteen industries. Their eighty-fur cnsumptin categries aggregate t ur fifteen cnsumer gds accrding t infrmatin in Appendix B f this chapter. We calculate the amunts f prducer gds used by the saving cmmdity (cmmdity 16) by using data n business investment frm the 1972 input-utput table, scaled up t 1973 levels. We impse the requirement that the ttal saving f cnsumers shuld exactly equal the net investment f the business sectr. After we add these amunts f prducer gds used by the saving cmmdity, we have a 19 x 16 matrix. This matrix gives the amunt f each industry's utput used in prducing each f the cnsumer gds. We divide each element f this matrix by its clumn ttal in rder t btain the Z matrix cefficients. These cefficients are presented in table This table is fairly straightfrward. The reader will nt be surprised t see that mst f fd is prduced by the fd industry, mst f husing by real estate, etc. Each cnsumer gd price is a weighted average f the prducer gd prices, where the weights are the elements f the apprpriate clumn frm the Z matrix. Demands fr cnsumer gds are calculated using the sixteen prices and the cnsumer demand functins. When we feed the vectr f ttal demands fr cnsumer gds thrugh the Z matrix, we get the cnsumptin demands fr the nineteen prducer gds, as shwn in figure Gvernment Enterprises Sme publicly supplied gds and services are subject t a price, r user charge. In ur mdel these gds and services are prduced by the gvernment enterprises industry (industry 19). We separate this industry frm the ther functins f gvernment. In this sectin we present data fr the gvernment enterprises industry. The data cver labr use and taxes, capital use and taxes, and subsidies received. Data n emplyee cmpensatin fr federal, state, and lcal gvernment enterprises are available in unpublished, disaggregated tables that crrespnd t table 6.1 f the July 1976 SCB. We present these data in

25 '*t-<tp 1 < -3 O0 CO 00 s (D ON NO NO IT) NO ON CO Tt fn NO * NO r CN > in ON NO r O g ON 00 ON NO in in I-- 00 V) < * O 8 ON g 55 U u 73 c,g cat G c 5 a (N r m NO CO NO NO CO OOOOOOOOOOOOOOOOONOO ''''' q p SSI ^H c in ON O O inrj- 5 M O 03 c in >n NO CN CO in pp^rppp ''' O O O O CO O O4 in O O O O O CO ' ' ' ' ' ' ' a ' 5 «3 60 c 'S <c l-i r* r- 3S3g >n s-g S s 3 u <u ^. J3 is S.a ««> "s e -e -a aj 03 *OJ ^ 5 u "S a ^ s

26 c 2 «ass <O O O inter < <u, 3 n 2 5 Z -a Z C 6C C <u f ancia vices c "i 5 O O VOOQO rf M O O O O O O O - ^ O O O O O O T t O O O O O * c c as CN in TJr- m 00 T-H CO VO qqqqqqhqnqqqq^qqq ^ q '' qqqqqqqqqqqqqqqqqqq '''i-j' c in 'O Tf ^j" r^- c t~- as CN CN CN O t-~ c i i \ >n c qqqqqqqqqqqqqqq ''''' S3 s > H 5 qqqqqqqqqqqqqqqqq<-h 'd'd duc g OH "B O c u ssifi 0 a-s.s P- a S-S 6 ^ -c a 1 c^ ff-si c8 3 <U > 9- t- j>-jl 81 &5 r^ i i II

27 80 Chapter Fur 19 x 16 Z matrix f cefficients fr 19 industries and 16 cnsumer gds 16 x 1 vectr f cnsumptin levels fr 16 cnsumer gds 19 x 1 vectr f derived cnsumptin demands fr the 19 prducer gds Fig. 4.2 Use f the Z matrix t create a vectr f derived cnsumptin demands fr the prducer gds. table Emplyer and emplyee cntributins t retirement prgrams are given in table 3.11 f the July 1976 SCB. The ttal cntributin fr gvernment emplyment is multiplied by the fractin f gvernment labr in gvernment enterprises, t prduce an estimate f the labr tax attributable t this industry. The ttal tax is $1,557 millin. When we subtract this figure frm the grss-f-tax wage bill f $16,723 millin, we get a net-f-tax labr bill f $15,166 millin. The rate f tax n net-f-tax labr is Capital estimates are cnsiderably mre difficult t prduce. The SCB shws a prfit-type return fr gvernment enterprises. This crrespnds t their "surplus," which is ften negative. Even if the aveiage surplus were psitive, this wuld nt be an apprpriate measure f the return t capital in gvernment enterprises. Our prcedure is t impute the amunt f capital, used by the gvernment enterprises industry, n the basis f capital /labr ratis in thse private activities that are mst similar t the activities f the gvernment enterprises. Seven cmpnents f this industry are listed in table 4.12 alng with the cmparable private industries. Since emplyee cmpensatin data were Table 4.11 Gvernment Enterprise Labr Use and Tax (in millins f 1973 dllars) Federal State and Lcal Ttal 1. Gvernment enterprise labr 2. Ttal gvernment labr 3. Prprtin f labr gvernment enterprises (1-^2) 4. Emplyer retirement cntributins 5. Emplyee retirement cntributins 6. Ttal cntributins (3 + 4) 7. Estimated labr tax in gvernment enterprises (6 x 3) $10,484 $62, $3,217 $2,246 $5,463 $6,239 $103, $6,828 $3,765 $10,593 $16,723 $165, $10,045 $6,011 $16,056 $918 $639 $1,557

28 T3 O a O tn _ *3 3 O u U T3 S 2 "3 ^3 8 & PH *3 O O w u PH WO

29 82 Chapter Fur nt available in this detail, and since value added includes the mismeasured surplus, we used the sum f intermediate inputs as a measure f the relative size f these activities. The inputs shwn in clumn (1) are frm the 1967 detailed input-utput table. We assume that the prprtins in clumn (2) wuld still be apprpriate in Clumn (4) cntains the capital/labr rati (grss-f-tax) fr the private cunterparts f these activities, btained frm industry data presented earlier in this chapter. When we use grss-f-tax ratis, we impute a higher capital/labr rati fr the gvernment enterprises industry than might seem apprpriate, because the gvernment pays n tax n the capital it uses. Hwever, we believe this is apprpriate treatment, since it als implies the same rati f the marginal prduct f capital t the marginal prduct f labr in bth the public and private prtins f each industry. We use the relative sizes f the seven cmpnents t get weighted averages f these capital/labr ratis fr federal and fr state and lcal gvernment enterprises. Multiplying the federal rati f.4020 by the federal gvernment enterprises' use f labr capital f $10,484 millin yields a capital estimate f $4,125 millin. The state and lcal labr rati is Multiplying this by labr f $6,239 millin yields a $3,596 millin capital estimate. Their sum f $7,811 millin is the final grss-f-tax and net-f-tax capital estimate fr gvernment enterprises. Finally, we use an estimate f the subsidy t the gvernment enterprises industry. Capital use is valued at $7,811 millin, while the 1973 surplus f gvernment enterprises (frm tables 3.2 and 3.4 f the July SCB) ttals nly $1,601 millin. The difference f $6,210 millin is the implied subsidy n these peratins. NID wrksheets list a single indirect business tax f $78 millin fr this industry, s the final utput tax attributed t it is - $6,132 millin. When we divide this figure by the ttal demand fr the utput f the gvernment enterprises industry ($22,564 millin), we have a 27 percent rate f subsidy n the utput f this industry. The effect f this negative utput tax is t lwer the utput price faced by purchasers. Purchases by gvernment ther than thse fr public enterprises are cnsidered an element f final demand and are described in the next chapter. 4.8 Other Prductin Taxes The NID has prvided a wrksheet with their estimates f the amunt f each indirect business tax paid by each industry in Each f these taxes requires a mdel-equivalent treatment that matches as clsely as pssible the legal base, rates, and rules f the tax. State and lcal sales taxes are represented by varius rates f tax n cnsumptin. We will discuss these taxes in chapter 5. The mtr vehicles