View
1
Download
0
Category
Preview:
Citation preview
AN ECONOMETRIC MODELOF UNEMPLOYMENTINSURANCE TAX RECEIPTS
A CBO Technical Analysis Paper
August 1977
Congress of the United StatesCongressional Budget OfficeWashington, D.C.
AN ECONOMETRIC MODEL OF UNEMPLOYMENT
INSURANCE TAX RECEIPTS
The Congress of the United StatesCongressional Budget Office
For sale by the Superintendent of Documents, U.S. Government t rinting OfficeWashington, D.C. 20402
PREFACE
This technical staff paper is one of a series that des-cribes the methodology for deriving estimates on a variety ofbudget issues and related topics. Such papers are intended toaid persons engaged in the more technical aspects of budgetaryand related questions of public policy. In accordance with theCongressional Budget Office's mandate to provide objective andimpartial analysis, this paper contains no recommendations.
The author of this paper is Marc P. Freiman. The authoracknowledges the helpful comments of Robert Black of CBO, JamesManning and Joseph Mickey of the Unemployment Insurance Service,and Ed Morrison of the Office of Management and Budget. Thanksalso go to Roger Winsby of Data Resources, Inc. for his technicalassistance, and to Mel Visnick of the Department of Treasury forproviding the revenue data. The manuscript was edited by DavidHowell Jones. Betty Ingram and Jill Bury typed the severaldrafts.
Alice M. RivlinAugust 1977 Director
m
TABLE OF CONTENTS
Page
Preface iii
CHAPTER I. Summary and Introduction ... 1
CHAPTER II. The Model of the State Funds
CHAPTER III. FUTA Taxes and Railroad Retirement 19
APPENDIX 23
TEXT TABLES
1. Total Taxable Wages in Private Industry(Billions of Dollars) 11
2. Average State Tax Rates 14
3. State Taxes, 1968 through 1975 15
4. Estimated FUTA Taxes, 1967 through 1976 20
APPENDIX TABLES
1. Average, Maximum and Minimum State Tax Rates, 1976 . 24
2. Total Covered Payroll In Private Industry ...... 26
3. Total Covered Employment in Private Industry .... 27
4. States with Taxable Wage Bases Greater thanthe FUTA Base 28
5. Trust Fund Earnings 29
vn
CHAPTER I. SUMMARY AND INTRODUCTION
This paper is a description of a model of unemploymentinsurance tax receipts which was developed at the CongressionalBudget Office. The model is used to estimate expected futurereceipts and to estimate the changes in revenues which wouldresult from legislative initiatives.
The unemployment insurance system is financed by a two-part payroll tax levied at the state and federal levels. The taxapplies to almost every employer in covered industries. Theprincipal industries which are not now covered are state andlocal government, agriculture, and domestic service. Mostemployers in these industries will be covered beginning in 1978.
The taxable payroll for each covered employer is definedunder federal law as total wages paid, up to a limit of $4,200per employee for each year. This $4,200 limit is called the"taxable wage base." Twenty-two states have instituted highertaxable wage bases for their state taxes. On January 1, 1978,the federal taxable wage base will rise to $6,000.
The Federal Tax
The Federal Unemployment Tax Act (FUTA) established a taxof 3.2 percent on taxable payrolls. The tax is reduced by 2.7percent, however, for employers in states with approved unemploy-ment compensation programs. Since all states have federallyapproved programs, an employer's effective federal tax rate isonly 0.5 percent. These FUTA taxes flow into three federalaccounts:
• the Employment Security Administration Account (ESAA),which finances federal and state costs associated withadministering the unemployment compensation program;
94-147 O - 77 - 2
• the Extended Unemployment Compensation Account (EUCA),which pays for 50 percent of the benefits under theExtended Benefits (EB) program. Up until March 31,1977, this account also financed the Federal Supple-mental Benefits (FSB) program, which is now paid fordirectly out of general federal funds; and
• the Federal Unemployment Account (FUA), which providesinterest-free loans to states that cannot meet theirbenefit costs. If the loans are not repaid after agiven period, the law requires that the federal tax onemployers in the state be increased until they arerepaid (unless the Secretary of Labor grants a waiverof this penalty increase).
Beginning January 1, 1977, the effective FUTA tax rate wasraised to 0.7 percent and will remain there until all loans tothe EUCA account have been repaid.
The State Tax
The taxes levied at the state level go into separate unem-ployment insurance trust funds for each state. I/ In any givenstate these funds are used to pay for the regular benefits andhalf the extended benefits drawn by unemployed persons whopreviously worked in that state.
While the federal tax remains at a uniform rate of 0.5percent, all 50 states and the District of Columbia have enactedan experience-rating system. This varies an employer's tax rateon the basis of the benefits received by his employees, therebypenalizing employers whose employees experience high levels ofunemployment by imposing higher tax rates on their taxablepayrolls (up to a maximum tax rate established by each state).Hawaii, Washington, and the District of Columbia have temporarilysuspended their experience-rating systems in favor of a uniformrate, and Puerto Rico has no provision for experience ratingunder the Federal Unemployment Tax Act (FUTA).
I/ In addition to the 50 states, the District of Columbia andPuerto Rico are also treated as separate jurisdictions in theunemployment insurance system.
Under federal law, states may have minimum tax rates aslow as zero. At present, four states are using minimum taxrates this low. Maximum tax rates can rise above 2.7 percent,and three states are now applying a tax rate of at least 6.0percent to some employers. The average tax rate for calendar1976 is estimated by the Department of Labor to be approximately2.5 percent. The maximum, minimum, and estimated average taxrates for each state in calendar 1976 are presented in AppendixTable A-l.
A Note on Modeling Tax Receipts and Sources of Data
The primary purpose of this model is to estimate UI taxreceipts in future periods as a function of economic assumptionsgenerated by the CBO. Because these assumptions do not generallycontain disaggregations of economic series, the type of modelingwhich can be performed is constrained. In particular, variousstructural approaches are for the most part not feasible.
For example, covered employment is estimated from employ-ment in the entire economy. This is not very revealing from astructural point of view. Covered employment could, on a his-torical basis, be estimated as a function of employment trendsin specific industries.
Economic downassumptions of the CBO do not, however, con-tain detailed breakdowns of employment by industry. If break-downs of employment by industry were used in the historicalestimation of UI taxes, therefore, then a second model would beneeded to estimate these industry breakdowns as a function oftotal employment for future periods. Besides entailing substan-tial extra effort, this indirect approach would probably notachieve significantly better results. Hence, this model wasconstructed so that the only independent variables used are thoseregularly available in the economic assumptions of the CBO.
The source of the data on unemployment insurance revenuesfor this model is the form BA-R 1114 of the Division of Govern-ment Financial Operations, Department of the Treasury. This formis issued quarterly and is therefore more convenient to use thanthe Monthly Treasury Statement (MTS). The form BA-R 1114also contains information such as interest earned, trust fundbalances, and advances, including state-by-state detail, none ofwhich is available in the MTS.
Because of the structural changes that occur over the longrun in the economy, and because of the difficulties involved inconstructing a complete data set, it was decided somewhat arbi-trarily to start the estimation period in 1965. Because thequarterly III revenue data exhibit seasonal fluctuations andthe economic assumptions of the CBO are seasonally adjusted, theequations which follow incorporate quarterly seasonal dummies.T-statisties are printed in parentheses beneath the coefficients.
CHAPTER II. THE MODEL OF THE STATE FUNDS
Basic Structure of the Model of State Taxes
The state tax portion of the unemployment insurance taxreceipt model can be described briefly as follows:
First, total covered wages are estimated from total wagesand salaries in the entire economy. Then, covered employmentis estimated from total employment. These two variables yield anaverage annual wage in covered employment.
The ratio of total taxable payroll to total covered payrollis then estimated as a function of the ratio of the taxable wagebase to the average covered wage and also as a function of time.Multiplying the ratio of total taxable payroll to total coveredpayroll by the total covered payroll yields estimated totaltaxable payroll.
Next, the national average state tax rate is estimated frompast ratios of state trust fund reserves to taxable payrolls.Multiplying total taxable wages by this estimated average statetax rate yields estimated state tax revenues.
Earned interest for a given period is estimated as afunction of the balance in the trust fund, the interest rate, andthe difference between revenues and outlays. This interest isthen added to taxes to yield total state trust fund revenues.
Finally, the initial trust fund balance for the next periodis calculated from the balance, taxes, interest, and benefitpayments during the last period.
This system can be described symbolically:
total covered wages = f(total wages + salaries in economy)
covered employment = f(total employment in economy)
average covered wage -
f/taxable wage base , tl-me\\average covered wage )
total taxable wagestotal covered wages
total covered wages x = total taxable wages
average tax rate - ffcff ^bl enroll
total state U I tax receipts = total taxable wages x averagetax rate
earned interest = f( previous balance, interest rate,revenues-outlays)
total state trust fund receipts = state taxes + interest
balance t+, = balance t + taxest + interestt - benefit payments^
Total Covered Wages
Total covered wages in private industry are primarily afunction of wages and salaries in the entire economy for the sameperiod. In addition, a dummy variable is added to reflect thechange in the coverage of the U I system in 1972. I/
log(covered wages) = -1.58 + 0.969 log (wages+salaries)(-19.1) (72.0)
+ 0.069 D-Cov - 0.029 Season2(10.7) (-7.1)
+ 0.038 Seasons +0.71 Season4(9.0) (17.9)
R2 = .999 DW = 2.058 Rho = 0.16
I/ The firm size requirement for employers in covered industrieswas lowered from eight or more workers to one worker (onat least one day in each of 20 weeks in a calendar year).The logged relationship used reflects the assumption thatthe independent variables wi l l have a constant proportionalr e l a t i o n s h i p wi th the dependent v a r i a b l e , rather than aconstant absolute relationship.
wherecovered wages = total covered wages per calendar
quarter (billions of dollars)
wages and salaries = quarterly wages and salaries(billions of dollars) in entireto 1 from 1972 on, zeroelsewhere
D-Cov = dummy for extension of coverage ofunemployment compensation system,set equal to 1 from 1972 on, zeroelsewhere
SeasonZ, Seasons, Season4 = dummies for second, third, and fourthcalendar quarters, respectively
The actual values for covered wages, the predicted values,and the difference between the two are displayed in AppendixTable A-2.
Covered Employment
Covered employment is estimated as a function of totalemployment in the economy, and once again a dummy is included toreflect the extension of coverage of the unemployment insurancesystem. 2/
log(covered employment) = -4.64 + 1.37 log (employment)(-5.82) (19.38)
+ 0.063 D-Cov + 0.025 Season2(9.44) (12.58)
+ 0.039 Season3 + 0.028 Season4(17.38) (13.90)
R2 = .997 DW = 1.74 Rho = 0.74
2/ The logged relationship used again reflects the assumptionthat the independent variables will have a constant propor-tional relationship with the dependent variables, rather thana constant absolute relationship.
where
covered employment = average covered employment in privateindustry per quarter (thousands).
employment = average employment in entire economy,seasonally adjusted (thousands).
D-Cov = dummy for extension of coverage of UCsystem, set equal to 1 from 1972 on,zero before 1972.
Season2, Seasons, Season4 = dummies for second, third, and fourthcalendar quarters respectively.
The coefficient of 1.37 on the employment variable indi-cates that covered employment has grown at a slightly fasterpercentage rate than total employment, even after controlling forthe extension of coverage in 1972. This is probably a reflectionof the fact that coverage of additional employment has beenelected at the state level on a continuing basis over time, notjust as a result of federal requirements such as the legislationwhich took effect in 1972. As coverage approaches universality,this coefficient should decrease in future reestimations of thisequation.
Appendix Table A-3 presentsand errors for covered employment.
the actual values, estimates,
Average Covered Wage
This variable is not estimated from exogenous variables,but is calculated by dividing estimated covered wages by esti-mated covered employment.
Ratio of Taxable Wages to Covered Wages
In order to derive total taxable wages, the ratio of totaltaxable wages to total covered wages is estimated as a functionof the ratio of the taxable wage base to the average coveredwage, p lus a time trend.
This specification was chosen for a variety of reasons.The r e l a t i o n s h i p between the ratios was the one be l i eved tobe the most stable. With the inclusion of the taxable wage basein the rat io , t h i s e q u a t i o n a lso a l l o w s one to est imate theeffects of changes in this important parameter of the program.Finally, the time trend serves to capture trends in the dis-persion of earnings and possibly in turnover of employees whichwould affect the relat ionship between the ratios but are noteasily modeled. 3_/
taxable wages 1.310 (AW])+ 0.873 lAW2)+ 0.511 \AW^)+total wages ~ (107.4)v '' (70.6) v V (43.8P *'
0.364 lAWrJ- 0.00151 T(29 .7) X 4/ (-5.39)
R2 = .995 DW = 2.12
where AW = the average quarterly wage per coveredworker
WB = taxable wage base
3/ The square-root relationship was chosen because it is aneasily performed transformation which embodies the expectedrelationship between the ratio of taxable to total wages andthe ratio of the wage base to average wages—namely thattaxable wages = 0 when the wage base = 0 and thattotal wagestaxable wages increases as the wage base increases,total wagesbut at a decreasing rate. This relationship can also bedescribed as one where the first derivative is positive andthe second derivative is negative.
94-147 O - 77 - !
WBj^r = one fourth of the taxable wage base,divided by the average covered wagein a given quarter
WB_i> WB WBq, WB, _ WBAW' AW, AW:: AWT " AW for the first, second,
1 * J H third, and fourth calen-dar quarters respec-tively, equal to zerootherwise. 4/
T = time trend, set equal to 1 in 1965:111.
The taxable wage base variable (WB) is not the FUTA wagebase because at any given time various states will have adoptedtaxable wage bases higher than the FUTA base. Instead, anaverage "effective" wage base was constructed by weighting eachstate's wage base by its proportion of covered wages. Theresulting series and the source data are presented in AppendixTable A-4.
Total Taxable Wages
Total wages are obtained by multiplying estimated totalcovered wages by the ratio described in the preceding section.These estimates are presented in Table 1.
The advantages of estimating UI receipts on a quarterlybasis become apparent here. Because of the low taxable wagebase, unemployment insurance tax receipts are strongly seasonal.For example, the total of $6.4 billion in state taxes for fiscalyear 1976 was distributed 23 percent, 14 percent, 11 percent,and 52 percent among the four quarters. As is demonstrated inTable 1, the model captures this seasonality in taxable wages.
The quarterly model is also useful in estimating the ef-ects of program changes because increases in the taxable wagebase usually occur at the beginning of a calendar year, butrevenue estimates are also needed on a fiscal year basis.
4/ Alternatively, these four variables can be described asthe result of taking four seasonal dummies and multiplyingeach one by WB.
AW
10
TABLE 1. TOTAL TAXABLE WAGES IN PRIVATE INDUSTRY: FOR CALENDAR YEARS 1965(FIRST QUARTER) THROUGH 1975 (FIRST QUARTER); IN BILLIONS OF DOLLARS
Year
1965: IV
1966:11966:111966:1111966: IV
1966
1967:11967:111967:1111967:IV
1967
1968:11968:111968:1111968: IV
1968
1969:11969:111969:1111969: IV
1969
1970:11970:111970:1111970: IV
1970
1971:11971:111971:1111971:IV
1971
1972:11972:111972:1111972:IV
1972
1973:11973:111973:1111973:IV
1973
1974:11974:111974:1111974: IV
1974
Actual
18.0
61.247.728.519.3
156.7
66.648.127.518.9
161.1
72.050.428.720.3
171.4
78.252.429.721.2
181.5
82.552.028.519.4
182.4
82.851.928.220.0
182.9
98.969.240.327.9
236.3
108.772.442.630.8
254.5
117.674.643.030.2
265.4
Estimated
18.1
65.345.126.819.5
156.7
68.146.727.619.7
162.1
71.449.529.120.5
170.5
75.652.130.521.0
179.2
80.053.330.320.5
184.1
81.354.230.621.0
187.1
102.569.839.527.3
239.1
109.074.241.828.3
253.3
115.277.543.128.2
264.0
Error
+0.1
+4.1-2.6-1.7+0.2
+1.5-1.4+0.1+0.8
-0.6-0.9+0.4+0.2
-2.6-0.3+0.8-0.2
-2.5+1.3+1.8+1.1
-1.5+2.3+2.4+1.0
+3.6+0.6-0.8-0.6
+0.3+1.8-0.8-2.5
-2.4+2.9+0.1-2.0
0.0
+1.0
+0.9
-2.3
+1.7
+4.2
+2.8
-1.2
-1.4
1975:1 120.7 116.4 -4.3
Although the estimates are certainly acceptable on a quar-terly basis, it should be noted that the percentage error for anyannual period is generally even less because of offsetting errorsin the quarterly estimates.
Average State Tax Rate
As noted earlier, each state has its own system for deter-mining an employer's tax rate. These systems involve some typeof formula which relates an employer's record in laying offworkers to the size of his taxable payroll. An experienceschedule yields a specific tax rate for a specific value of theformula. Alternate schedules may be used, depending on thecondition of the trust fund of the entire state.
There are basically four categories of formulas:
(1) reserve-ratio formulas, in which an employer's reservesare divided by his taxable payroll.
(2) benefit-ratio formulas, in which an employer's benefitpayments are divided by his taxable payroll.
(3) benefit-wage-ratio formulas, in which the taxable wagesof those workers who become unemployed and receivebenefits are divided by the employer's total taxablepayrol1.
(4) payroll variation formulas, in which the tax rate is afunction of the percentage change in an employer'spayroll over time.
The reserve-ratio formula is by far the most popular methodand is used by 33 states at present. These states contained 62percent of all covered workers and collected 69 percent of allstate taxes in calendar 1975.
In the following equation, the average state tax rate wasmodeled approximately along the lines of a reserve-ratio experi-ence-rating formula, using a second-degree distributed lagconstrained to equal zero at both ends:
12
average state tax rate. = 2.88 - 3.12 reserve ratio. 91 (53.5) (-23.3) W
- 4.99 reserve ratio. ., - 5.61 reserve ratio.(-23.3) *•-* (-23.3)
- 4.99 reserve ratio. - 3.12 reserve ratio(-23.3) (-23.3)
,.t-
R = .990 DW = 1.89 Rho = 0.57
where average state tax rate = the ratio of tax receipts to totaltaxable wages for the year endingin quarter t
reserve ratio = the ratio of the trust fundt-i balance at the start of quarter
t-i to total taxable wages for theyear ending with quarter t-i
The quarterly moving summations were used in order toretain a quarterly time frame yet eliminate the nonrelevantseasonal fluctuations in tax receipts. Regressions were run withvariables added to reflect other experience-rating formulas (forexample, a distributed lag on the ratio of benefits to taxablepayroll). These variables were not statistically significant,however, and did not add substantially to the explanatory powerof the equation. Experience rating in states which do not employa reserve-ratio formula apparently still approximates the resultswhich would have been obtained if such a formula had been used.
The actual average tax rates, the estimated rates, and theerrors are presented in Table 2. However, this tax rate equa-tion is not now in use by the Congressional Budget Office becauseof the current extreme condition of the trust fund. Some stateshave introduced special charges, thereby temporarily altering theexperience-rating schedules. Furthermore, many state trust fundbalances include non-interest-bearing advances which also distortthe normal functioning of the experience-rating system. Theeffective tax rate, therefore, is now estimated on an ad hocbasis after taking into account expected borrowing and otherdevelopments at the state level.
13
TABLE 2. AVERAGE STATE TAX RATES: CALENDAR YEARS 1968 (FIRSTQUARTER) THROUGH 1975 (FIRST QUARTER); IN PERCENTS
Year a/
1968:11968:111968:1111968: IV
1969:11969:111969:1111969:IV
1970:11970:111970:1111970: IV
1971:11971:111971:1111971:IV
1972:11972:111972:1111972:IV
1973:11973:111973:1111973:IV
1974:11974:111974:1111974:IV
1975:1
Actual
1.561.531.511.49
1.441.421.411.41
1.381.371.371.36
1.411.431.451.46
1.621.691.721.75
1.881.962.002.02
2.001.991.981.98
1.97
Estimated
1.581.531.501.48
1.471.441.431.41
1.411.391.381.38
1.391.431.471.50
1.541.661.741.79
1.851.952.012.02
2.022.001.991.98
1.99
Error
+.02.00
-.01-.01
+.03+.02+.02
.00
+.03+.02+.01+.02
-.02.00
+.02+.04
-.08-.03+.02+.04
-.03-.01+.01
.00
+.02+.01+.01
00
+.02
a/ For the twelve-month period ending with the quarter listed.
14
Total State Tax Receipts
Total state tax receipts are produced by multiplying esti-mated taxable wages by the estimated average state tax rate.Because there is approximately a one-quarter lag in receipt offunds by the treasury, tax receipts for a given quarter are theproduct of the taxable wages and tax rate for the precedingquarter. The estimates of state tax receipts are presented inTable 3, where they are compared with the actual figures reportedin the appropriate volumes of The Budget of the United States.
Included in the state tax figures are payments made to thetrust fund on a reimbursable basis. Some state and local gov-ernments and nonprofit organizations have elected to financeunemployment compensation benefit payments on this basis. In-stead of being taxed beforehand, these employers are "billed"whenever a former employee collects benefits. Such reimburse-ments have not always been consistently reported as trust fundincome by all states, although they are now required to do so.
TABLE 3. STATE TAXES: FISCAL YEARS 1968 THROUGH 1975
Year
1968
1969
1970
1971
1972
1973
1974
1975
EstimatedTaxableWages b/
(Billionsof Dollars)
165.4
174.7
183.6
185.4
208.3
245.6
259.5
265.2
EstimatedTaxRate b/
(Percents)
1.56
1.44
1.38
1.41
1.62
1.88
2.00
1.97
EstimatedStateTaxes b/
(Billionsof Dollars)
2.58
2.52
2.53
2.61
3.37
4.62
5.19
5.22
ActualStateTaxes a/
(Billionsof Dollars)
2.60
2.56
2.56
2.58
3.23
4.63
5.26
5.30
Error(Billions
of Dollars)
-0.02
-0.04
-0.03
+0.03
+0.14
-0.01
-0.07
-0.08
a/ SOURCE: The Budget of the United States Government, years 1970-1977.
b/ Because of the one-quarter lag in depositing receipts in the Treasury,these columns are calculated for the 12-month period ending with thefirst quarter of the calendar year.
15
These reimbursable payments introduce other inaccuraciesinto the estimation procedure. Workers in nonprofit organiza-tions are included in the employment figures in this model, butthe payments made on their behalf on a reimbursable basis are notidentical to those which would have been made through a payrolltax. Hence, estimates generated by applying a tax rate topayrolls will be slightly inaccurate. Covered state and localgovernment workers are not included in the employment figures butsome of the reimbursable payments for them are included in thestate tax figures. It is very difficult to model these paymentsexplicitly, but as long as they are constant or follow a trendthey will be implicitly incorporated into the estimates producedby the model.
When the coverage of state and local government employeesis significantly expanded, however (which will happen in 1978 asa result of PL. 94-566), adjustments must be made. As an approxi-mate solution to this situation, the CBO estimating procedureadds to the estimate of state taxes an amount equal to theestimated additional benefit payments resulting from the newcoverage.
Earnings
State trust fund earnings are estimated as a function ofthe trust fund balance at the beginning of the period, the inter-est rate, and the adequacy of current income to cover outgo, is/
log(earnings) = -6.96 + 1.09 log(balance, start of period)(-5.37) (13.48)
+ 0.567 log(bond rate) + 0.584 flow adequacy(4.38) (7.47)
R2 = .974 DW = 1.64 Rho = 0.85
5/ The logged relationship here results from the fact that ifthe general equation is earnings = rate x balance, thentaking the logs provides a linear relationship.
16
where bond rate = yield on Moody's AAA corporate bonds (inpercentage points)
flow adequacy = deposits - withdrawalsbalance, start of period
and earnings and balance are measured in thousands ofdollars.
The flow-adequacy variable is related to the amount of fundbalance which will be used in benefit payments. The three-monthTreasury bill rate was tried as a measure of the interest rate,but it did not perform as well as the bond rate.
The actual values, the estimated values, and the errorsare presented in Appendix Table A-5.
Total State Trust Fund Receipts
Total state trust fund receipts are obtained by addingearnings to taxes.
The balance at the beginning of the next period is calcu-lated according to the following formula:
balance.+, = balance. + taxes. + earnings.
- benefit payments.. 6/t
6/ For a description of the method used by CBO to estimatebenefit payments, see Estimating Outlays for UnemploymentCompensation Programs, Congressional Budget Office, TechnicalAnalysis Paper No. 1, October 27, 1976.
17
Beginning early in calendar 1975, a number of state trustfunds have required repayable advances which have totaledhundreds of millions of dollars in most of the quarters sincethen. These advances complicate the process of calculating theend-of-quarter balance. Although large amounts of new advanceswill probably not be needed in the foreseeable future, some statefunds will continue to borrow, and the repayment of these advan-ces should also be included in the calculations.
Sufficient data have not been available to create a rig-orous model of this advance-and-repayment process. At present,therefore, this process is dealt with on a quarter-to-quarter adhoc basis, taking into account new data and anticipated develop-ments. Although the new balance calculation is important inconsidering the status of the trust fund and projecting budgetauthority, it should be emphasized that it only affects a smallpart of direct trust fund receipts (i.e., earnings), and eventhis small part is not included in a unified budget.
18
CHAPTER III. FUTA TAXES AND RAILROAD RETIREMENT
FUTA Taxes
FUTA taxes are the taxes collected at the federal level.These taxes are estimated in a manner similar to that used toestimate state taxes.
Total covered wages, total covered employment, and averagecovered wage are estimated by means of the equations described inthe state model. The ratio of taxable wages to total wages isthen also calculated using the equation described in Part Two.The FUTA taxable wage base is used, however, instead of theaverage effective state wage base. !_/
FUTA taxes are then derived by multiplying the estimatedtotal taxable wages by the effective FUTA tax rate. This effec-tive tax rate was 0.4 percent from 1961 to 1969, was 0.5 percentfrom 1970 to 1976 (with a 0.08 percent surcharge added forcalendar year 1973), and was raised to 0.7 percent in 1977. Itwill remain at this level until all loans to the FUTA accountsare repaid.
The estimated FUTA taxes, along with the actual receiptsrecorded in the Budget of the United States, are presented inTable 4.
The FUTA base was $3,000 through 1971, and $4,200 forthe period 1972-1977, and it will be $6,000 in 1978. Thereare no aggregate data available on wages subject to the FUTAtax.
TABLE 4. ESTIMATED FUTA TAXES: FISCAL YEARS 1967 THROUGH 1976
Year
EstimatedTaxableWages
(Billions)
FUTATax
Rate
EstimatedFUTATaxes
(Billions
ActualFUTATaxes a/
(BillionsError
(Billionsof Dollars) (Percent) of Dollars) of Dollars) of Dolars)
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
151.2
157.0
164.9
172.4
174.6
201.9
243.2
256.7
261.9
266.5
0.4
0.4
0.4
0.4/0.5
0.5
0.5
0.5/0.58
0.58/0.5
0.5
0.5
0.605
0.628
0.660
0.765
0.873
1.01
1.30
1.40
1.31
1.33
0.589
0.601
0.629
0.770
0.964
1.01
1.30
1.45
1.36
1.53
+0.016
+0.027
+0.031
-0.005
-0.091
0
0
-0.05
-0.05
-0.20
I/ SOURCE: The Budget of The United States Government, variousyears.
Earnings and the new trust fund balance are calculated byusing the equations presented for the state trust fund. As isthe case with the state trust fund, the recent recession createda large need for advances to the federal extended unemploymentcompensation account (EUCA), which financed federal supplementalbenefits (FSB) and half the cost of extended benefits (EB). Themeager reserves in this account were exhausted during the secondquarter of 1975, and thereafter FSB and the federal share of EBwere in effect paid out of general revenues (by means of advan-ces). Because of the change to general revenue financing ofFSB and the increase in the FUTA tax rate, however, furtheradvances are not anticipated in the future.
20
Railroad Unemployment Tax Receipts
The financial transactions of the railroad unemploymentinsurance system are made through the unemployment trust fund.These transactions are small in relation to the other trustfund components, however. In fiscal year 1976 for example,railroad unemployment tax receipts were less than 2 percent thesize of state tax receipts.
Because of its small size this account was not modeled as atrust fund. Instead, estimated receipts were set equal toestimated outlays for administration and benefit payments. Theestimated outlays were generated by a separate model. 2/
Total Receipts from All Sources
Total receipts for the entire unemployment trust fund areobtained by combining the state taxes, the FUTA tax, railroadretirement contributions, and earnings from all sources.When revenues are needed on a unified budget basis, however,earnings are not included in the total, because the earnings areboth paid out and received by the federal government.
See Estimating Outlays for Unemployment Compensation Pro-grams, Congressional Budget Office, Technical AnalysisPaper No. 1, October 27, 1976.
21
APPENDIX
TABLE A-l. AVERAGE, MAXIMUM, AND MINIMUM STATE TAX RATES, 1976 a/
State
Al abamaAlaskaArizonaArkansasCalifornia
ColoradoConnecticutDel awareDistrict of ColumbiaFlorida
GeorgiaHawaiiIdahoIllinoisIndiana
IowaKansasKentuckyLouisianaMaine
MarylandMassachusettsMichiganMinnesotaMississippi
MissouriMontanaNebraskaNevadaNew Hampshire
MaximumPercentage
4.04.82.94.44.9
3.64.54.52.74.5
4.033.03.64.03.3
4.73.64.23.35.0
3.65.16.66.02.7
3.23.13.73.54.15
AveragePercentage(estimated)
1.83.71.91.83.6
1.93.02.52.72.1
1.83.01.71.91.8
2.32.32.51.93.1
2.04.13.71.82.2
2.82.22.63.22.5
MinimumPercentage
0.52.30.10.51.4
0.01.61.62.70.7
0.053.00.50.10.3
0.70.00.40.72.4
2.8 b/3.90.80.91.3
0.51.50.11.12.4
(continued)
24
TABLE A-l. (CONTINUED)
State Maximum Average (est.) MinimumPercent Percent Percent
New Jersey 6.2 3.4 1.2New Mexico 3.6 1.9 0.6New York 5.2 3.5 1.5North Carolina 4.7 1.4 0.3North Dakota 4.2 2.2 0.9
Ohio 4.3 2.3 0.6Oklahoma 2.7 1.7 1.2Oregon 4.0 3.3 2.6Pennsylvania 4.0 2.9 1.0Puerto Rico c/ 3.45 3.0 2.95Rhode Island 5.0 3.9 3.2
South Carolina 4.1 2.1 1.3South Dakota 2.7 1.0 0.0Tennesse 4.0 1.6 0.4Texas 4.0 0.6 0.1Utah 2.8 1.7 1.3
Vermont 5.0 2.3 1.0Virginia 2.7 1.2 0.55Washington 3.0 3.0 3.0West Virginia 3.3 1.9 0.0Wisconsin 5.2 2.1 0.5Wyoming 3.86 2.2 1.16
United States 2.5
a/ SOURCE: Significant Provisions of State Unemployment Insur-ance Laws, January 3, 1977, and Attachment to Un-employment Insurance Program Letter 28-76, bothU.S. Department of Labor, Unemployment InsuranceService.
b/ The disparity between the minimum and the average tax ratesfor Maryland results from the fact that Maryland raised itsminimum rate from 0.7 percent to 2.8 percent in the middle ofthe year.
£/ Puerto Rico does not have an experience-rating system. Thisrate applies to employers subject to Puerto Rico law, but notto the FUTA.
25
TABLE A-2. TOTAL COVERED PAYROLL IN PRIVATE INDUSTRY: IN BILLIONSOF DOLLARS
Year
1965: IV
1966:11966:111966:1111966: IV
1967:11967:111967:1111967: IV
1968:11968:111968:1111968: IV
1969:11969:111969:1111969: IV
1970:11970:111970:1111970: IV
1971:11971:111971:1111971:IV
1972:11972:111972:1111972: IV
1973:11973:111973:1111973: IV
1974:11974:111.974:1111974: IV
Actual
70.3
65.469.773.075.6
71.574.176.180.0
77.781.283.488.7
85.289.692.197.8
90.894.696.6
100.7
93.699.9
103.3108.4
110.4115.4118.5125.0
123.5129.3132.4139.9
135.0141.7145.9152.1
Estimated
69.2
66.069.471.975.8
71.474.376.480.7
77.181.584.288.8
84.489.292.396.8
91.894.996.9
100.3
95.9100.0102.7108.1
111.3116.7119.7127.2
121.9128.9132.7140.7
133.0140.4145.0151.5
Error
-1.1
+0.6-0.3-1.1+0.2
-0.1+0.2-0.3+0.7
-0.6+0.3+0.8+0.1
-0.8-0.4+0.2-1.0
+1.0+0.3+0.3-0.4
+2.3+0.1-0.6-0.3
+0.9+1.3+1.2+2.2
-1.6-0.4+0.3+0.8
-2.0-1.3-0.9-0.6
1975:1 141.6 140.6 -1.0
26
TABLE A-3. TOTAL COVERED EMPLOYMENT IN PRIVATE INDUSTRY: IN MILLIONSOF DOLLARS
Year Actual Estimated Error
1965: IV
1966:11966:111966:1111966: IV
1967:11967:111967:1111967: IV
1968:11968:111968:1111968: IV
1969:11969:111969:1111969: IV
1970:11970:111970:1111970: IV
1971:11971:111971:1111971:IV
1972:11972:111972:1111972: IV
1973:11973:111973:1111973:IV
1974:11974:111974:1111974: IV
46.18
45.3947.3548.7648.61
47.1848.4049.6349.52
48.3849.9951.3451.29
50.3252.1653.4652.99
51.4552.2352.8651.55
50.6652.1853.1152.92
55.8858.2659.5060.10
59.5861.8162.9463.13
61.4363.2764.1563.00
46.03
45.3046.9148.3948.53
47.0548.6349.6649.51
48.0550.2650.8251.02
50.3151.8253.2153.05
51.6452.5253.0052.31
50.3652.2653.4853.26
55.9958.0559.6959.46
59.2261.8263.0163.07
61.7663.0764.2262.88
-0.15
-0.09-0.44-0.37-0.08
-0.13+0.23+0.03-0.01
-0.33+0.27-0.52-0.27
-0.01-0.32-0.25+0.06
+0.19+0.29+0.14+0.76
-0.30+0.08+0.37+0.34
+0.11-0.21+0.19-0.64
-0.36+0.01+0.07-0.06
+0.33-0.20+0.07-0.12
1975:1 59.52 60.11 +0.59
27
TABLE A-4. STATES
WITH
TAX
ABLE
WAG
E BASES
GREA
TER
THAN T
HE F
UTA
BASE:
CALE
NDAR
YEA
RS 1965 T
HROU
GH 1976
00
Alas
kaAr
izon
aCa
lifo
rnia
Conn
ecti
cut
Delaware
Hawa
iiId
aho
Massachusetts
Mich
igan
Minn
esot
aNe
vada
New Jersey
North
Dakota
Oregon
Penn
sylv
ania
Rhod
e Is
land
Tennessee
Utah
Verm
ont
West
Vir
gini
aWi
scon
sin
Wyomi ng
Alab
ama
Geor
gia
Iowa
Miss
ouri
Mont
ana
Puerto R
ico
Wash
ingt
on
Aver
age State
Taxa
ble
Wage
Base
a/
1965
7,200
3,600
3,800
3,600
4,200
3,600
3,600
3,600
3,800
3,600
3,600
3,600
3,300
4,200
3,600
3,600
1966
7,200
3,600
4,100
3,600
4,300
3,600
3,600
3,600
4,800
3,800
3,600
3,600
3,600
3,300
4,200
3,600
3,600
3,600
7 3 3 3 4 3 3 3 4 3 3 3 3 3 4 3 3 31967
,200
,600
,800
,600
,600
,600
,600
,600
,800
,800
,600
,600
,600
,300
,200
,600
,600
,600
1968
7,200
3,600
3,800
3,600
3,600
4,800
3,600
3,600
3,600
4,800
3,800
3,600
3,300
3,600
3,600
3,600
3,300
4,200
3,600
3,600
3,600
3,600
1969
7,200
3,600
3,800
3,600
3,600
5,000
3,600
3,600
3,600
4,800
3,800
3,600
3,400
3,600
3,600
3,600
3,300
4,200
3,600
3,600
3,600
3,600
1970
7,200
3,600
3,800
3,600
3,600
5,500
3,600
3,600
3,600
4,800
3,800
3,600
3,800
3,600
3,600
3,600
3,300
4,200
3,600
3,600
3,600
3,600
1971
1972 1973
7,200
7,200
7,200
3,600
3,800
3,600
3,600
6,000
6,300
6,500
3,600
3,600
3,600
4,800
4,800
4,800
3,800
3,600
4,000
4,400
3,600
3,600
3,600
3,600
4,200
3,600
3,600
3,600
3,600
1974
10,000 10 6
6,800
7
4,800
4 4
5,000
5 4
(3/4 year)4
All
1975
,000
,000
,300
,800
,800
,000
,800
,800
All
wages
wages
3,244
3,321
3
SOURCES:
Hand
book of
Une
mplo
ymen
tDepartment of
Labo
r,
,289
3,328
3,330
Insu
ranc
e Fi
nanc
ial
3,332
4,200
4,800
5,400
3,357
4,263
4,273
Data
, 1938-1970,
(Un
empl
oyme
nt1971)
and Jo
seph H
icke
y, U
nemp
lom̂ient
Insu
ranc
e Se
rvic
e
6,000
6
4,324
4
,600
,361
1976
10,000
7,000
6,000
7,800
7,800
5,400
6,200
6,100
5,400
7,000
4,800
6,000
6,000
4,800
6,000
6,000
4,500
4,800
All
wages
7,200
4,976
Insu
ranc
e Se
rvic
e,.
NOTE
: FU
TA
Tax
able
W
age
Base
was
$3,
000
thro
ugh
1971
and
$4,
200
the
rea
fte
r,
a/
Ave
rage
d
eri
ved
by w
eig
hti
ng
sta
tes
by t
he
ir
cove
red
pa
yro
lls
in 1
970.
TABLE A-5. TRUST FUND EARNINGS: IN MILLIONS OF DOLLARS
Year
1965: IV
1966:11966:111966:1111966: IV
1967:11967:111967:1111967:IV
1968:11968:111968:1111968: IV
1969:11969:111969:1111969: IV
1970:11970:111970:1111970: IV
1971:11971:111971:1111971:IV
1972:11972:111972:1111972: IV
1973:11973:111973:1111973: IV
1974:11974:111974:1111974: IV
1975:11975:111975:1111975:IV
Actual
72.8
72.576.987.592.4
93.396.8
103.1105.1
105.7108.6122.1124.0
123.8127.5141.1143.8
143.5146.8162.2157.2
144.6136.6125.9119.9
109.2107.0112.3113.8
109.6116.5144.9148.4
145.4150.7173.6162.9
131.9108.384.855.4
Estimated
75.0
74.478.688.093.8
90.097.5
106.7111.5
106.7109.1114.6128.0
129.3130.1137.6150.4
147.9147.4151.0155.0
136.8136.5134.2117.9
105.7105.9113.4113.9
109.2115.7132.7149.0
142.6148.6164.2166.7
126.996.585.261.9
29
O
Error
+ 2.2
+ 1.9+ 1.7+ 0.5+ 1.4
- 3.3+ 0.7+ 3.6+ 6.4
+ 1.0+ 0.5- 7.5+ 4.0
+ 5.5+ 2.6- 3.5+ 6.6
+ 4.4+ 0.6-11.2- 2.2
- 7.8- 0.1+ 8.3- 2.0
- 3.5- 1.1+ 1.1+ 0.1
- 0.4- 0.8-12.2+ 0.6
- 2.8- 2.1- 9.4+ 3.8
- 5.0-11.8+ 0.4+ 6.5
Recommended