Dedicated Revenue Mechanisms for Freight Transportation Investment (2012)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

70 Economic Impact Methodology The research team used existing econometric models to estimate the relative economic impacts of the leading revenue mechanism candidates. This analysis was undertaken to deter- mine the overall impact of a freight infrastructure tax or fee, any differences between the candidate mechanisms, and the relative impact on different transportation and industry seg- ments. As in the cost and efficiency comparisons, the analysis set the tax burden at $5 billion annually for all options to maintain comparability. The economic impacts of tax scenarios were estimated by “following the dollar” of tax, implementation, and compliance costs. For each of the leading revenue mechanism candidates, these costs are immediately felt at the vehicle level—either at the pump (the fuel tax scenarios), by mile traveled (VMT fee scenario), or by the vehicle (registration fee scenario). In the short term, for-hire trucking firms will bear some of these costs. Ultimately, the additional costs will be reflected in higher rates and higher delivered commodity prices, which will translate into higher industry production costs, and ultimately affect the growth potential in these industries. Along the way, truck- ing companies and industries with private fleets will shift their behavior to minimize their tax burden. The methodology for estimating the economic impacts of each scenario essentially follows the same six steps: • Estimate tax, implementation, and compliance costs. • Estimate how each cost accrues to trucking market segments. • Estimate how costs pass through market segments to yield higher commodity prices. • Estimate how higher commodity prices accrue to industries. • Estimate how higher production costs affect industry growth potential. • Estimate possible substitution effects as fleets reduce their tax burden. The six steps would apply to freight-carrying, Class 4–8 trucks in the following four major tax scenarios: • Diesel fuel tax with non-freight tax refunds. • Diesel/gas tax with vehicle ID. • VMT distance/vehicle fee—OBU/options. • Annual registration fee. Cost Accrual by Trucking Market Segments The research team determined how scenario tax, implemen- tation, and compliance costs would accrue to six trucking market segments: • Private (Fixed Contract) • Private (Variable Contract) • For-Hire (Long-Term Fixed) • For-Hire (Long-Term Variable Contract) • For-Hire (Short-Term Fixed Contract) • For-Hire (Short-Term Variable Contract) • For-Hire (Spot Market) The study team determined how the basis of taxation or cost accrual varies by trucking segment. Information was compiled from VIUS, the Freight Analysis Framework (FAF), the U.S. Census Bureau, and other sources to determine the share of total trucks by market segment. Truck counts by class and private/for-hire split were determined by scaling 2002 VIUS truck counts up to 2008 levels according to vehicle counts estimated by the U.S. Census. The resulting truck counts by class are shown in Table 40. Further allocation for private and for-hire truck counts into contract markets was accomplished by applying assumptions about average annual miles per truck to market segment mileage data. Average operating characteristics were determined by estimating how each commodity reported by FAF uses each C h a p t e r 6 Economic Impacts

71 of the six market segments. These utilization rates were com- bined with 2008 FAF data to determine the average VMT, commodity loading, backhaul percentage, and average haul length for each market segment. These results are summarized in Table 41. Cost Accrual by Commodity and Industry Once cost accrual by market segment was determined, it was possible to determine how costs ultimately passed through to commodities, resulting in higher commodity prices for industries consuming them. This pass-through would not be immediate. In for-hire market segments, some costs would be temporarily borne by the trucking sector, only to be passed through after contracts could adjust accordingly. Assumptions about the timeline for pass-through costs are as follows: • Diesel fuel tax. Costs begin accruing 1 year after implemen- tation begins. – Tax cost. In the first year of cost accrual, two-thirds of costs are passed through to consumers (with one-third absorbed by the for-hire trucking industry). In all sub- sequent years, the tax cost is fully passed through. – Implementation and compliance cost. In the first and second year of cost accrual, 1/3 and 2/3 of costs are passed through to customers, respectively. In all subsequent years, costs are fully passed through. • Diesel/gas fuel tax and VMT fee. Costs begin accruing 5 years after implementation begins. – Tax cost. In the first year of cost accrual, two-thirds of costs are passed through to consumers. In all subsequent years, the tax cost is fully passed through. – Implementation and compliance cost. In the first and second year of cost accrual, one-third and two-thirds of costs are passed through to customers, respectively. In all subsequent years, costs are fully passed through. • Registration fee. Costs begin accruing 5 years after implementation. – Tax, implementation, and compliance costs. In the first and second years of cost accrual, one-third and two-thirds of costs are passed through to consumers, respectively. In all subsequent years, the tax cost is fully passed through. The purpose of these assumptions is to acknowledge the differences in implementation timelines for low-tech mecha- nisms such as diesel tax refunds or registration fees, compared to technology-dependent diesel tax or VMT fee systems. Table 40. Estimated 2008 vehicle counts for freight-carrying trucks. Market Engine Type Class 4-5 Class 6 Class 7-8 Total Diesel 625,035 619,128 2,511,127 3,755,291 Gasoline/Other 769,852 452,090 251,924 1,473,866 Diesel 114,662 119,291 1,258,169 1,492,122 Gasoline/Other 141,228 87,107 126,223 354,559 1,650,777 1,277,616 4,147,443 7,075,838 Private For-Hire Totals Source: Tioga Group. Ton-trips moved by segment (millions) Total VMT including backhauls (millions) Ton-miles moved by segment (millions) Private (Fixed) 2,339 46,269 436,525 Private (Variable) 2,516 43,850 423,488 Private - TOTAL 4,856 90,119 860,013 Long-Term Fixed 795 18,351 165,436 Long-Term Variable 1,644 38,780 339,679 Short-Term Fixed 1,466 35,269 329,964 Short-Term Variable 4,539 78,033 737,140 Spot Market 724 13,465 124,750 For Hire - TOTAL 9,168 183,898 1,696,970 All Truck - TOTAL 14,024 274,017 2,556,983 Market Segment SEGMENT VOLUMES Average length of haul (miles) Backhaul % of VMT Avg. tons per loaded trip 187 31% 12.96 168 31% 13.2 208 30% 12.44 207 27% 12.11 225 28% 12.57 162 32% 12.97 172 33% 12.5 SHIPPING CHARACTERISTICS Sources: FAF, VIUS, Oak Ridge. Table 41. Baseline trucking market volumes and shipping characteristics—2008.

72 Actually, implementation timelines are not yet predictable given the conceptual state of most proposals. Estimated costs were passed through each market segment to translate to higher delivered commodity prices. For private market segments, costs were passed to commodities based on the unit of taxation or fee by commodity. For for-hire market segments, costs were passed to commodities assuming that the costs would be reflected in higher per-vehicle-mile shipping rates. The research team then determined which industries would ultimately bear the long-run costs of taxation, implementation, and compliance. The allocation was based on how industries use the affected commodities. Industries purchase either raw or intermediate commodities, add value, and sell the final product. As the delivered prices of any single commodity increase, industry impacts depend on how much of the final good value is determined by the affected commodity. This information is expressed as the “absorption,” or “use,” Table in an input-output economic model. Cost Effect on Industry Growth Potential For each of the industries listed in Table 42, the total costs of taxation, implementation, and compliance passed through from trucking translate into higher production costs. At the national level, these higher production costs reduce profits, thereby limiting the growth potential of industries by reduc- ing the amount of capital they have for investment. Lower rates of investment ultimately reduce long-run productivity growth rates. Therefore, to estimate this effect, a statistical relationship between industry profits and productivity was established using Bureau of Economic Analysis data on components of gross domestic product (GDP). With the statistical relationship between industry profits and productivity established, the study team then assumed that all production cost increases translated into lower industry profits and estimated the ensuing reduction in productivity. Results, shown in Table 42, indicate that a $5 billion tax burden reduces economic activity by roughly $500 million, so each additional dollar of truck tax or fee reduces long-run GDP by about $0.10. Associated with this lost GDP is lower personal income and general tax (corporate and personal income tax) revenue. Tax Substitution Effects Additional taxes or fees will lead businesses to change their behavior to minimize their tax burden. The research team ana- lyzed two potential strategies that businesses might use to mini- mize their tax burden if additional taxes or fees are imposed: • Substituting gasoline-powered trucks for diesel-powered trucks (applies to the diesel tax scenario only). • Shipping by rail intermodal instead of truck (applies to all scenarios). These substitution strategies were modeled because they were determined to be the primary options available to mini- mize tax burden. However, the following behaviors would also be expected to reduce net tax revenue, albeit to a lesser degree than the two listed above: • Higher load factors (applies to fuel and VMT tax), thereby shipping more ton-miles using fewer truck miles. • Better vehicle utilization (applies to excise tax), thereby shipping more freight using fewer trucks. • Fewer empty backhauls (applies to fuel and VMT tax), thereby shipping more ton-miles using fewer truck miles. Fleet Mix (Applies to Diesel Tax Only) For the diesel fuel tax, the most significant option businesses would have to minimize the tax burden is to switch to trucks that burn other fuels. To estimate this effect, the study team designed a simple model of the tradeoffs businesses face in Diesel Ta x Diesel/Gas Ta x VM T Fe e Registration Fe e Agriculture and Forestr y 3 3 3 2 Mining 50 53 59 45 Utilities 0 0 0 Construction 80 86 96 80 Manuf ac turing 147 155 171 129 W holesale Tr ade 24 27 30 30 Retail Tr ade 14 15 17 15 Tr ansportation and Wa rehousing 27 31 35 40 Services 121 132 147 133 Gover nm ent 0 0 0 0 0 TOTA LS 466 501 558 475 Scenario A gg re g ated Industr y Sector Table 42. Economic impact of increased taxes (GDP, $m2010).

73 deciding whether to buy a diesel or gasoline truck. The research team did not model costs for other fuel types, such as natural gas, assuming that the portion of the truck market using other fuel types would continue to be negligible for the forecast period in question. Generally, diesel-burning trucks are more expensive to purchase, but they deliver lower per-mile operating costs through better fuel economy, lower main- tenance costs, and longer operational life. Separate models were built for Class 4 and 5, Class 6, and Class 7 and 8 trucks. Model assumptions are shown in Table 43. This is a highly simplified model, suitable only for these high-level estimates. In Table 43, the key result is the break-even point—the annual miles per truck at which diesel and gasoline trucks cost approximately the same. When the fleet average miles per truck is above this threshold, the majority of trucks should be diesel; when the fleet average miles per truck is below the threshold, the majority of trucks should be gasoline. In reality, the fleet for each class is composed of trucks operated under a range of conditions. Table 44 shows how the fleet breaks down for private versus for-hire operation. Underlying these averages is a distribution of annual miles per truck, per truck class. The probability of any single operator purchasing a diesel truck rather than a gasoline truck will depend on how heavily the operator plans on utilizing the truck. To model the fleet mix change, a binomial logistical dis- tribution was assumed for each truck class type, where the probability of purchasing a diesel truck is based on the relative cost of diesel versus gasoline for that class’s average annual miles per truck. Once this distribution was fitted, the addi- tional tax cost was used to modify the relative cost and a new fleet mix resulted. Table 45 shows the model results. Diversion to Rail (Applies to Fuel and VMT Tax Scenarios) To estimate diversion to rail intermodal, the research team calibrated a binary mode choice for each commodity model using the FAF database. The mode choice increment was 10 tons of freight (an average truck load), and applied only to trip lengths greater than 700 miles. The intermediate and final results are shown in Table 46. Table 46 shows the wide variation in cross-modal elasticity by commodity, from a low of 0.0% for live animals to a high of 17.9% for plastics and rubber. In general, lower value com- modities with no special handling characteristics are more likely to switch to rail if the cost of trucking rises. Many such Table 44. Average miles per truck for private and for-hire fleet mix. Measure Class 4-5 Class 6 Class 7-8 Baseline truck count (2008) 1,650,778 1,277,617 4,147,443 Baseline diesel trucks 739,697 738,420 3,769,296 Baseline diesel truck percent 44.8% 57.8% 90.9% Baseline add'l cost of operating gas truck (c/mi.) 2.6 6.2 11.7 Baseline break-even annual miles/truck 11,400 19,370 25,580 New add'l cost of operating gas truck (c/mi) 0.3 3.3 8.9 New break-even annual miles/truck large ~30,000 ~40,000 New fleet % diesel 28.6% 45.5% 82.3% New diesel truck count 472,765 580,740 3,411,799 Lost diesel trucks 266,932 157,680 357,497 Avg. annual miles per lost truck 15,000 14,000 20,000 Lost truck-miles 4,004,486,488 2,187,558,646 7,151,258,814 Lost diesel gallons consumed 400,949,836 269,237,987 851,340,335 Lost tax revenue from fleet change 93,751,612$ 62,954,248$ 199,063,627$ Table 45. Fleet mix change modeling. *Source: VIUS, updated to 2008 using Census. Table 43. Diesel/gasoline cost model assumptions.

74 SCTG Code SCTG Description Annual Truck Trips >700 miles Cross Mode Elasticity Baseline Trucking Cost $m Diesel Tax Diesel/ Gas Tax 01 Live animals/fish 11,669 0.0% 2,710 22 22 02 Cereal grains 502,725 1.5% 28,522 290 310 03 Other ag prods. 1,494,025 3.7% 19,193 187 193 04 Animal feed 762,507 2.8% 10,856 158 185 05 Meat/seafood 1,302,373 3.0% 5,793 72 80 06 Milled grain prods. 2,316,159 3.6% 12,044 159 180 07 Other foodstuffs 4,133,421 3.0% 23,598 291 325 08 Alcoholic beverages 584,917 2.7% 4,723 50 52 09 Tobacco prods. 16,111 0.4% 142 1 2 10 Building stone 47,643 0% 668 8 9 11 Natural sands 271,546 0% 12,765 152 170 12 Gravel 701,350 0% 36,820 384 402 13 Nonmetallic minerals 563,918 0% 5,588 51 53 14 Metallic ores 10,341 0% 657 7 8 15 Coal 28,016 0% 3,742 29 27 16 Crude petroleum 0 0 % 2,035 16 15 17 Gasoline 182,339 0% 9,208 130 152 18 Fuel oils 361,242 0% 10,164 147 173 19 Coal-n.e.c. 733,348 0% 16,130 226 265 20 Basic chemicals 862,236 17.0% 6,524 53 51 21 Pharmaceuticals 492,108 8.4% 2,626 24 23 22 Fertilizers 1,598,636 10.1% 13,061 156 175 23 Chemical prods. 1,822,323 6.0% 17,124 207 232 24 Plastics/rubber 2,680,035 17.9% 13,328 120 119 25 Logs 31,519 0.1% 10,943 80 76 26 Wood prods. 3,168,079 4.3% 23,934 248 265 27 Newsprint/paper 1,200,216 17.1% 6,344 59 58 28 Paper articles 765,835 4.4% 5,331 59 64 29 Printed prods. 2,839,760 6.4% 11,974 118 119 30 Textiles/leather 1,040,118 8.6% 5,059 51 51 31 Nonmetal min. prods. 2,620,674 1.2% 25,080 310 346 32 Base metals 2,647,706 8.3% 19,495 178 176 33 Articles-base metal 2,391,032 3.7% 14,403 134 133 34 Machinery 1,107,983 1.9% 10,035 104 107 35 Electronics 1,472,081 9.1% 7,422 68 68 36 Motorized vehicles 2,188,905 8.4% 12,600 115 115 37 Transport equip. 77,837 6.7% 799 8 8 38 Precision instrument 312,082 10.4% 2,099 22 24 39 Furniture 1,261,080 4.0% 6,397 72 77 40 Misc. mfg. prods. 1,647,085 5.3% 9,130 99 109 41 Waste/scrap 290,032 0.7% 23,391 204 203 42 Unknown freight 1,143,072 13.7% 12,248 135 149 43 Mixed freight 105,593 0.6% 9,375 115 127 TOTALS 47,789,677 474,082 5,119 5,495 Baseline Data Add'l Shippin g Lost Trucks VMT Fee Reg. Fee Diesel Tax Diesel/ Gas Tax VMT Fee Reg. Fee Diesel Tax Diesel/ Gas Tax VMT Fee Reg. Fee 23 11 0 0 0 0 0.0 0.0 0.0 0.0 345 286 7 8 9 193 0.3 0.2 0.2 0.1 212 128 53 55 60 962 2.0 1.5 1.7 0.7 212 585 31 37 42 3,055 1.2 1.0 1.2 2.2 90 119 49 55 62 2,138 1.9 1.5 1.7 1.5 204 337 110 125 141 6,124 4.1 3.5 4.0 4.3 366 483 153 171 192 6,686 5.8 4.8 5.4 4.7 58 27 17 18 19 238 0.6 0.5 0.5 0.2 2 1 0 0 0 1 0.0 0.0 0.0 0.0 10 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 192 174 0 0 0 0 0 0 0 444 230 0 0 0 0 0 0 0 59 28 0 0 0 0 0 0 0 9 6 0 0 0 0 0 0 0 29 15 0 0 0 0 0 0 0 16 8 0 0 0 0 0 0 0 173 200 0 0 0 0 0 0 0 198 247 0 0 0 0 0 0 0 302 381 0 0 0 0 0 0 0 55 25 118 115 124 1,496 4.5 3.2 3.5 1.1 25 12 38 36 39 479 1.4 1.0 1.1 0.3 197 164 192 215 243 5,328 7.2 6.0 6.8 3.8 261 217 133 148 167 3,651 5.0 4.1 4.7 2.6 129 57 431 428 465 5,374 16.3 12.0 13.0 3.8 81 41 0 0 0 0 0 0 0 294 155 140 150 166 2,304 5.3 4.2 4.7 1.6 63 28 190 188 204 2,373 7.2 5.3 5.7 1.7 71 41 38 40 45 687 1.4 1.1 1.3 0.5 130 61 180 182 198 2,440 6.8 5.1 5.6 1.7 56 26 90 91 99 1,217 3.4 2.5 2.8 0.9 390 321 38 42 48 1,033 1.4 1.2 1.3 0.7 191 85 200 199 216 2,518 7.6 5.6 6.0 1.8 144 64 82 82 88 1,036 3.1 2.3 2.5 0.7 118 60 22 23 25 331 0.8 0.6 0.7 0.2 73 34 123 122 132 1,595 4.6 3.4 3.7 1.1 124 55 169 168 182 2,131 6.4 4.7 5.1 1.5 8 4 5 5 5 65 0.2 0.1 0.2 0.0 27 14 35 37 41 576 1.3 1.0 1.1 0.4 86 49 56 61 68 1,016 2.1 1.7 1.9 0.7 123 83 95 104 117 2,079 3.6 2.9 3.3 1.5 220 97 2 2 2 21 0.1 0.0 0.1 0.0 167 112 173 190 212 3,764 6.5 5.3 5.9 2.7 142 99 1 1 1 17 0.0 0.0 0.0 0.0 6,118 5,177 2,971 3,096 3,414 60,930 112.0 86.6 95.5 43.1 Cost ($m/yr) Lost Truck Miles (Billions) Lost Revenue ($m/yr) Table 46. Potential diversion to rail intermodal. Table 47. Long-run tax and cost accrual to industry groups ($m2010). commodities, however, are already moved predominantly by rail, so the potential for further modal shift is reduced. Economic Impact Findings The economic impact methodology described above yields a high-level understanding of how the different federal revenue mechanisms may impose costs on different industries in the U.S. economy over the long term and how those costs will affect the output of those industries. Since the reduced capacity for output is the principal economic impact on U.S. industries, economic impacts are summarized in terms of (1) costs imposed by indus- try and (2) output lost by industry. Further impact on different economic indicators could be derived from changes in output. Table 47 shows how tax, implementation, and compli- ance costs ultimately accrue to broad industry groups. On

75 a percentage basis, the fuel tax and VMT fee scenarios are all relatively similar. These three scenarios primarily differ in magnitude from implementation and compliance costs. The registration fee scenario, however, shows a somewhat different impact on industry sectors because the mechanisms of cost accrual are different for that single scenario. For the fuel tax and VMT fee scenarios, tax cost accrual to market segments is by truck miles (for the diesel tax, just diesel-truck miles). In contrast, the registration fee option accrues to market segments by truck count. This different mechanism of accrual translates into a significant difference in which market segments bear the immediate tax burden. Fundamentally, for-hire fleets use trucks much more inten- sively than private fleets. 2002 VIUS data (see Table 48) indicate that for-hire Class 6 through 8 trucks drive over twice as many miles per year as equivalent trucks in private fleets, and for-hire Class 4 through 5 trucks drive nearly twice as many miles per year as equivalent trucks in private fleets. This means that taxes accruing on a per-mile basis (the fuel and VMT options) more heavily affect for-hire fleets, whereas the registration fee, which is levied on a vehicle basis, accrues more heavily to private fleets. These costs are then passed on to commodities based on which commodities depend on market segments, and then to industries based on how they use commodities. Ultimately, the cost accrual pat- tern shown in Table 47, and the resulting economic impacts shown in Table 49, are the consequence of how costs first pass through market segments. Table 49 shows how the additional tax scenarios translate into long-run economic loss for the United States. With less profit from higher production costs, businesses will have less money to invest in product development and equipment, lead- ing to lower economic growth potential than would other wise occur. How these impacts occur by industry depends, first, on the magnitude of the increased production cost shown in Table 47. Beyond this, industries differ in their ability to turn investment into productivity. For example, the process of GDP growth for utilities and government sectors is relatively independent of production costs, and therefore higher pro- Source: 2002 Vehicle Inventory and Use Study (U.S. Census Bureau, 2004). Table 48. Average annual miles per truck by truck class and fleet type. *Does not include net tex revenue Table 49. Economic impacts of tax scenarios as lost 2021 GDP ($m2012). Diesel TaxRevenue Type Diesel/Gas Tax Gross Tax Revenue 5,000.0 5,000.0 Government Collection & Enforcement Costs -4.8 -70.8 Lost Revenue from Diesel to Gas Fleet Conversion -355.8 0.0 Lost Revenue from Truck to Rail Diversion -112.0 -86.6 Revenue, Net of Gov't Cost and Substitution Effects 4,527.4 4,842.6 VMT Fee Reg. Fee 5,000.0 5,000.0 -297.7 -50.0 0.0 0.0 -95.5 -43.1 4,606.8 4,906.9 Table 50. Net annual government revenues ($m2010).

76 duction costs have minimal impact on these sectors. In con- trast, manufacturing and service sectors are very sensitive to production costs and can readily convert increased profit- ability into investment and ensuing productivity. It should be noted that no long-run employment impacts are estimated to occur. Rather, the primary economic consequence of the four options (from industry’s perspective) is slower GDP growth and lower incomes. Finally, Table 50 shows net annual government rev- enues. The registration fee is the most efficient mechanism because it has low collection and enforcement costs, and it has minimal substitution effects. The diesel tax and VMT fee are both relatively inefficient compared to the other two scenarios. For the diesel tax, many businesses will be flex- ible in their choice of engine type and therefore migrate toward gasoline-powered trucks. For the VMT fee, the primary inefficiency is the relatively high collection and enforcement costs that come from essentially creating a new collection mechanism rather than expanding existing collection methods.