The Economic Impact of the Proposed Gasoline Tax Cut In Connecticut

Transcription

1 The Economic Impac of he Proposed Gasoline Tax Cu In Connecicu By Hemana Shresha, Research Assisan Bobur Alimov, Research Assisan Sanley McMillen, Manager, Research Projecs June 21, 2000 CONNECTICUT CENTER FOR ECONOMIC ANALYSIS Deparmen of Economics, U-63 Universiy of Connecicu Sorrs, CT Phone: (860) Fax: (860)

2 Inroducion This analysis esimaes he effec of a decrease in he Connecicu gasoline excise ax financed by an increase in he Sae income ax on he Connecicu economy using he single-region (saewide) REMI inpu-oupu model. The REMI model is a sophisicaed 53-secor replicaion of he sae s economic srucure, capable of projecing he economic impacs of various shocks up o he year Our objecive is o measure he long run economic impac of he gasoline ax cu and personal income ax increase on he economy in erms of several key economic variables, including oal employmen, personal income and Gross Sae Produc (GSP). The analysis looks a he impac over a period of eleven years ( ). Assumpions and Mehodology This analysis examines several quesions: 1. Wha is he impac of a gasoline ax cu on he sae budge? Wha will be he loss in sae ax revenue as a resul of a gasoline ax cu in he unbalance budge case? 2. If we assume a balanced budge, wha will be he size of he sae income ax increase necessary o offse he loss in sae revenue? Wha will be he long-run economic consequences of he gasoline ax cu and offseing income ax increase on he whole sae? 3. Wha will be he long-run economic impac in he case of he unbalanced budge? 4. Finally, is i worh i for he sae o reduce he gasoline ax? We esimaed he impac of he gasoline ax cu on he sae budge based on a proposed 7 cens per gallon ax cu. Appendix 1 presens he daa used in he model. 2

3 Appendix 2 presens a deailed descripion of our economeric esimaions. The model suggesed he own price elasiciy of gasoline consumpion was , and he income elasiciy of gasoline consumpion was Using hese esimaes we calculaed a projeced increase in gasoline consumpion afer he ax cu, which hen allowed us o find ha he projeced ne loss in gas ax revenue o he sae is $86.9 million. Under he model of a balanced budge, we propose ha he loss in he revenue from gasoline ax cu is offse by an equivalen increase in personal income axes. We model he reducion in he gasoline ax by adjusing he Consumer Expendiure Price Index in REMI by an equivalen dollar amoun. We offse his loss of ax revenue by an increase in income ax by he same amoun. Because he average marginal federal income ax rae is 29% 1, he sae income ax increase implies an increase in personal income ax of $61.7 million (71% of $86.9 million). In effec, he Federal governmen subsidizes sae ax increases. To see he long-run effecs of he gasoline ax cu under he balanced budge scenario, we employed he REMI model using he $86.9 million reducion in gasoline prices and $61.7 million increase in personal income axes derived earlier in he repor. Table 1 in Appendix 3 shows he resuls. Under he firs model of an unbalanced budge, we capure he gasoline ax cu by reducing governmen expendiures on highways by $86.9 million and adjusing he Consumer Expendiure Price Index in REMI by an equivalen dollar amoun. 2 Table 2 in Appendix 3 shows he resuls of his model. 1 See All Tax Cus are No Equal by Fred V. Carsensen, The Connecicu Economy, Spring 1999, p We use revenues from he gasoline axes for highway mainenance. 3

4 The second model of an unbalanced budge assumes ha he gasoline ax cu will be offse by inernal reallocaion of funds and sreching ou some paymens. In his scenario we adjus he Consumer Expendiure Price Index in REMI by $86.9 million wihou any oher changes in he budge. Table 3 in he Appendix 3 shows he resuls. In each case above he gas ax cu induces income and subsiuion effecs because relaive prices change. We expec subsiuion of Connecicu gas for nearby saes gas, because here are no maerial subsiues for gasoline. In he balanced budge case, he posiive income effec of he gas price decrease maches somewha he decrease in disposable income. In he unbalanced budge case, he income effec probably resuls in somewha increased gas consumpion perhaps manifesed in SUV proliferaion. Resuls The increase in he sae income ax required o offse he proposed 7 cens per gallon gasoline ax cu can be achieved by an increase of he highes marginal ax rae from 4.5% o %, which accouns for approximaely $53 in addiional ax per year for an average axpayer in Connecicu earning $62,000 adjused gross income. Assuming a conservaive 20,000 miles per year for an average driver and 22 mpg for an average car, we ge savings of approximaely $64 from he gasoline ax cu, which leaves an average axpayer almos indifferen o he proposed change. 3 The ables in Appendix 3 summarize he REMI simulaion resuls. The ables presen five variables ha measure he economic effecs of impacs examined: gross sae produc (GSP), oal employmen, populaion, personal income, and real disposable 3 For our purposes an average driver is an average axpayer. 4

5 personal income. The ables show an average of he annual increases (or decreases) of hese aggregae levels ha flow direcly and indirecly from he proposed gasoline ax cu and income ax increase over eleven years, compared o he baseline (saus quo) forecas of he Connecicu economy s performance. For example, he value for GSP in he second column of Table 1 indicaes ha, on average, here will be an increase in GSP of $ million per year over he REMI model s baseline forecas in he case of a 7 cens per gallon gasoline ax cu offse by an increase in sae personal income axes over 11 years. The resuls of our analysis show ha alhough he proposed 7 cens per gallon gasoline ax cu may leave an average axpayer indifferen, he sae economy will definiely benefi from i. Our resuls are relaively conservaive and slighly underesimae rue economic impac on he economy, as hey do no ake ino consideraion he capure of addiional gasoline consumers resuling from he decrease in Connecicu s relaive gasoline prices compared o our hree neighboring saes (New York, Massachuses, Rhode Island). We assume ha he fundamenal gasoline price is he same in he four saes and differences are due only o sae axes. The oal sae ax in each sae is: 29.8 cpg for New York, 21.5 cpg for Massachuses, 28 cpg for Rhode Island, compared o 35.3 cpg (32 cpg excise ax and 3.3 cpg 5% gross earnings ax colleced a wholesale) for Connecicu before he cu. Afer he reducion in Connecicu, he relaive price is only subsanially differen for Massachuses. We assume ha Connecicu residens formerly going o New York or Rhode Island o buy gasoline will now be indifferen and probably buy heir gasoline in Connecicu. People who formerly bough heir gas in Massachuses will likely coninue o do so because 5

6 here is sill some benefi o do so and habiual behavior is slow o change wihou subsanial incenive. We assume ha people in souhwesern Connecicu did no ravel o New York o buy gas, because of heir high opporuniy cos, and herefore he cu will no change heir behavior much. We esimae ha wealhy people will benefi less han poorer people will gain from his proposed policy change. 6

7 Appendix 1 Daa on gasoline consumpion, gasoline prices and income per capia in he Sae of Connecicu Year Consumpion of gasoline Price of gasoline Income per capia Jul ,388, , Aug ,152, , Sep ,786, , Oc ,883, , Nov ,397, , Dec ,924, , Jan-94 98,650, , Feb-94 91,796, , Mar ,829, , Apr ,432, , May ,780, , Jun ,774, , Jul ,092, , Aug ,791, , Sep ,315, , Oc ,838, , Nov ,479, , Dec ,338, , Jan ,639, , Feb-95 99,300, , Mar ,447, , Apr ,195, , May ,321, , Jun ,342, , Jul ,261, , Aug ,559, , Sep ,190, , Oc ,873, , Nov ,518, , Dec ,537, , Jan-96 99,616, , Feb-96 97,617, , Mar ,542, , Apr ,657, , May ,627, , Jun ,436, , Noe: The daa has been deflaed by he Connecicu CPI. Source: The daa on gasoline consumpion was provided by he Sae of Connecicu Deparmen of Revenues Services. The daa on gasoline prices was provided by he Harford office of AAA. The daa on income per capia was generaed using he daa from he Survey of Curren Business by Bureau of Economic Analysis. Year Consumpion of gasoline Price of gasoline Income per capia Jul ,065, , Aug ,339, , Sep ,446, , Oc ,605, , Nov ,987, , Dec ,761, , Jan ,045, , Feb-97 90,527, , Mar ,563, , Apr ,637, , May ,823, , Jun ,670, , Jul ,812, , Aug ,357, , Sep ,999, , Oc ,111, , Nov ,177, , Dec ,564, , Jan ,068, , Feb-98 92,837, , Mar ,242, , Apr ,006, , May ,608, , Jun ,571, , Jul ,759, , Aug ,555, , Sep ,202, , Oc ,569, , Nov ,242, , Dec ,044, , Jan ,353, , Feb-99 98,512, , Mar ,970, , Apr ,693, , May ,401, , Jun ,428, ,

8 Appendix 2: Esimaion of he change in he sae gasoline ax revenues: Par 1. Esimaion of price elasiciy and income elasiciy of gasoline consumpion. We consruced he following model for gasoline consumpion: ln C = α + α ln P where υ = ρυ C I α ln I and ε is iid(0, 2 σ ε = deseasonalized consumpion of gasoline = + ε deseasonalized real income per capia 2 + υ P = deseasonalized real price of gasoline The model gives he following resuls: Ep = price elasiciy of gasoline consumpion = (: ) Ei = income elasiciy of gasoline consumpion = (: 2.21) R 2 = 0.663, DW = 1.64, r = ) Par 2. Esimaion of he change in consumpion of gasoline as a resul of a gasoline ax cu. P 1999 = average price for fiscal year = P = Then he percenage change in gasoline price is % P = P/P 1999 = Then we find he percenage change in gasoline consumpion as a resul of a change in gasoline price: % Cp = Ep * % P = Now we find he prediced change in gasoline consumpion as a resul of a direc change in gasoline price: Cp = % Cp * C 1999 = 37,293,676 8

9 where C 1999 = cumulaive gasoline consumpion for fiscal year = 1,380,735,892 The examinaion of income per capia series suggess ha hey have a linear rend wih he slope of Therefore, he prediced percenage change in income per capia is % I = Then we find he percenage change in gasoline consumpion as a resul of a prediced change in income per capia: % Ci = Ei * % I = Now we find he prediced change in gasoline consumpion as a resul of a prediced change in income per capia: Ci = % Ci * C 1999 = 1,667,208 Then he oal prediced change in gasoline consumpion is: C = Cp + Ci = 38,960,884 Par 3. Esimaion of he change in sae gasoline ax revenues. TR = C τ C τ = C τ + C Therefore, TR = 38,960,884 * ,380,735,892 * (-0.07) = -$86,911,291 τ 9

10 Appendix 3: REMI Simulaion Resuls Table 1. Annual average changes in seleced economic variables from 7cpg decrease in gasoline ax financed by an increase in personal sae income ax ( ). Annual Average Employmen (Unis) 739 GRP (Mil. 92$) Personal Income (Mil. Nominal $) Real Disposable Personal Income (Mil. 92$) Populaion (Unis) 1104 Table 2. Annual average changes in seleced economic variables from 7cpg decrease in gasoline ax under he non-balanced sae budge assumpion wih a decrease in governmen expendiures ( ). Annual Average Employmen (Unis) 581 GRP (Mil. 92$) Personal Income (Mil. Nominal $) Real Disposable Personal Income (Mil. 92$) Populaion (Unis)

11 Table 3. Annual average changes in seleced economic variables from 7cpg decrease in gasoline ax under he non-balanced sae budge assumpion wihou a decrease in governmen expendiures ( ). Annual Average Employmen (Unis) 1,531 GRP (Mil. 92$) Personal Income (Mil. Nominal $) Real Disposable Personal Income (Mil. 92$) Populaion (Unis)