New User Special Price Expires in

Let's log you in.

Sign in with Facebook


Don't have a StudySoup account? Create one here!


Create a StudySoup account

Be part of our community, it's free to join!

Sign up with Facebook


Create your account
By creating an account you agree to StudySoup's terms and conditions and privacy policy

Already have a StudySoup account? Login here

Public Economics

by: Madie Schinner

Public Economics ECN 230A

Madie Schinner
GPA 3.57


Almost Ready


These notes were just uploaded, and will be ready to view shortly.

Purchase these notes here, or revisit this page.

Either way, we'll remind you when they're ready :)

Preview These Notes for FREE

Get a free preview of these Notes, just enter your email below.

Unlock Preview
Unlock Preview

Preview these materials now for free

Why put in your email? Get access to more of this material and other relevant free materials for your school

View Preview

About this Document

Class Notes
25 ?




Popular in Course

Popular in Economcs

This 108 page Class Notes was uploaded by Madie Schinner on Tuesday September 8, 2015. The Class Notes belongs to ECN 230A at University of California - Davis taught by Staff in Fall. Since its upload, it has received 76 views. For similar materials see /class/191888/ecn-230a-university-of-california-davis in Economcs at University of California - Davis.


Reviews for Public Economics


Report this Material


What is Karma?


Karma is the currency of StudySoup.

You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!

Date Created: 09/08/15
LECTURE HEALTH INSURANCE AND LABOR MARKETS HILARY HOYNES UC DAVIS EC230 OUTLINE OF LECTURE 11ntr0ducti0n and background 2 Theory of health insurance and mobility Madrian Job Lock QJE 3 Theory of health insurance and wages 4 Gruber AER Mandated Maternity Benefits Health Insurance and Labor Markets Introduction 0 United States is distinctive in that most insurance is provided through employers 0 Typically employer plans are group plans in nature 0 90 of private insurance is through employer own or spouse Implications for Labor Market 0 Large fraction of total compensation consists of employeeemployer premiums 0 Up 300 over 30 years ConcernsIssues Declining job growth Declining international competitiveness Labor market inefficiencies J oblock mobility Declining wages for workers Premium increases passed through as lower wages Literature Focuses on impacts on mobility earnings employment hours worked Insutuuonal Background of Health Insurance and Labor Markets Facts about nonelderly health insurance Dec 39ning employer provided insumnce Increasing Medicaid usagecoverage Increasing of uninsured um x959 mm mm um I991 199 I994 ms mum Polluuuun mm 100 a um me 0 me a mu 0 mu 0 Im A Inn H mm mm 14 9 u x 73 I 71 l m g m 1 7o 1 m m I er 5 5 m m 55 7 as n m s s 4 31 a 8 MN m 13 323 30 m 305 m m m 354 m 5 m m no ix mo 3H 0er mm s s 5 s 53 i s s u s 9 n 7 71 59 mm mm 3 2 u n m 3 I st loo 6 7 m m Ummum t ISJ m 15R mu m m m m m Source Gruber Health Insurance and Labor Markets Handbook of H with Economics Role of Workplace Or Why is so much insurance provided through workplace Pooling Economics 0 Share fixed high administrative costs 0 Mitigate effects of adverse selection although it is still a problem for small firms 2 Tax a ea uctibilily of premiums for employers Cause or effect Employer often covers 100 of H1 costs 3 Antidiscrimination Regulations 0 Illegal to offer HI selectively to highly compensated employees 0 Cannot offer only to some employees and not others 0 This is important for theoretical significance All mer m w I M 1 w suss W249 m t 1mqu en Empln m mm m men m Fmplnvm Fmplnvcc Fmplh a om Insurance ms m m o m u 951 Family Cover mm D m u 222 M77 om was a 942 a 9m CovenL1 by Insurance 1 m n 74 n m u m n mu a 717 n 525 Taken m 7 ms 11749 0 7n n m o m a 792 a as w No ImumM mum J m u 313 m in am mm u 459 0mg l ovenge 1m 0 m u u I n m 0 am a 77 a 397 I39m cum Insl um Weakly Enmmg 5259 no 3 m l m 2 su l w 1 5m x m omnm 3 755 0501 u m ms om o m u m mm nnm Sr Dmhlhl 1 7n n s3 0 ms 0 m an H w 3490 a 3m 0420 om 0 SB n has Wuuy Earnings 252 9 2m 2 252 s m i m 5 4x 9 1m 9 Funny mm Pnnsmn 1 mg n m 0 mm u m o m o m u ans an orrcnsT nmmuy 7 me n 1221 n m n m u m o m n m m cm 077 n mu 0077 a m Chancteristics of EPHI Gmbmmalm Handbook of l I Coverage decreases with rm size no rm size taker Variation I Reasons cited for no insurance Large rms Ineligible Preexisting condition Waiting period 0 mall rms Covered elsewhere Fam39 7 Note Preexisting Conditionimeans plan will not cover costs of illnesses existing before enrollment during a Waiting period or never Theory of Mobility Simple Compensating wage dz erentials model Assume I H1 is a 01 dummy for covered or not covered I Homogeneous HI plan available everywhere I Perfect experience rating 9 can perfectly price discriminate at worker level I Utility of job j for individual i depends on wages and HI L10 2 uwlj H U I Continuum of j obs supplieddemanded under completive conditions I HI costs for each person constant across firms Ci 2 CI for all j Workers I Suppose the worker considers some reduction in wages Awa in return for obtaining the homogeneous HI I They will desire HI if compensating wage differential is not too big I Will choose lower wage and health insurance over higher wage and no HI if utility is higher uwlj AwiPl uwlj0 gt O I Define this difference as Vi F irms I Firm j will offer HI to worker i if Awij gt C1 I Competition will bid down wage until Awij C l 1 This implies that wi Awij WU C l Therefore the wage falls by the full cost of the insurance Impact Ole on Labor Market in Simple Model I Workers who value insurance at 2 Cl will choose HI and get lower wages incurring 100 of cost I If worker has Vi gt 0 valuation of H1 above cost then the worker receives economics rents I No allocative inefficiency as workers find best job match regardless of H1 How does our economy di er from this Simple model 1 Employers cannot set employee specific compensation packages Must offer HI to all workers or none at all Administrative costs absorb rents 2 Costs of insurance varies across firms 9 Ci 72 Cle Implications I Matching 9 workers who value HI will select into firms that offer HI I Firms who can provide HI cheaply will do so I Workers will work at firm if Vi gt 0 I Firms will offer HI if C J lt Aw Job Lock Does the model explain job lock I Suppose a worker is at job 0 with wio and would be more productive at job 1 where wt1 gt wio I But further suppose that the HI costs are higher at firm 1 C1 gt C0 0 Perhaps due to a worse experience rating I Result Firm 1 decides not to offer HI C1 too high Firm 1 knows it could attract worker i to firm if it could offer HI but it would have to offer HI to all workers legally and that would be too expensive I Therefore the worker s choice to stay in the job 0 IF 11in Awl uwl10 gt 0 then the worker will stay in job 0 I Inefficiency 9they would be more productive if they move but they do not Firm 0 could extract rent knowing that worker i locked in But can they discriminate in this way not likely This result is not unique to H1 It is generally true if I Workers have different valuations of employer benefits I There are differential costs of provision across employers I There is an inability to set workerspecific compensation packages Other examples in HI I Non homogenous HI range of quality in plans I Locked out of retirement If MU gt MP worker should retire but leisure labor may be locked in due to reliance on EPHI Health Insurance and Mobilitv Empirical Evidence Job to Job Mobility o 20 million Americans change jobs each year 0 12 million leave jobs with EPHI These 12 million people have 7 million dependents o Potentially millions more who don t leave for fear of losing health insurance or facing limitations on coverage at new jobs 67 of EPHI plans have preexisting condition clauses Waiting periods for these conditions can be from 6 months 2 yrs 0 What is the impact of health insurance on mobility decisions Job Lock Anecdotal Evidence Surveys suggest that 1130 of individuals report that they or a family member remained in a job because they didn t want to lose health insurance 20 of those who reported being locked attribute it to preexisting conditions Goal Can survey evidence be confirmed in a real context for mobility decisions Empirical Issues 1 Compare mobility from jobs with vs without HI We would expect lower mobility from jobs with HI Problem Selection issues on both worker and firm side Worker Less healthy workers choose firms with HI Health may be correlated with mobility May overestimate mobility effects F z39rm Firms with HI not comparable to those without Fact Workers in firms with HI have much higher earnings and much higher pension benefits Good Jobs vs Bad Jobs Difficult to disentangle other impacts on mobility that are correlated with HI may be indiVidual or family characteristics 2 Group Comparison Approach 1990s literature Find two groups for Whom job lock should operate more strongly for one than the other Look at effects of EPHI on mobility across these 2 groups Di erence in Di erence EPHI No Yes Value of H1 High M 00 M 01 Low M 10 M11 Expect M11 M01 gt0 mobility of those With low valuation of H1 should be higher than those with high valuation of H1 But M 10 Moo Will capture the difference in mobility rates that eXist for across those With high vs low valuation of H1 DinD Mll M01 M10Moo Emplovment Based Health Insurance and Job Mobilitv Is there Evidence of Job Lock Madrian QJE Note Most prominent example of this approach Reasons for Job Lock 1Preexisting conditions 6 months 2 years waiting list 2 Length of service requirements for eligibility for benefit 3Discrimination in hiring of those employees with high perceived costs particularly in small firms O settmg Factors 1 COBRA legislation 2Employer must pay 100 of employee s premium for 18 months Approach Compare highlow risk groups Compare those WithWithout employer provided HI Difference in Difference Problem Risk not always observed Data from labor market surveys has mobility info but not HI info Data from medical surveys has HI info but not mobility Alternative Definition of Risk With vs Without alternative HI spouse Those W insurance less impacted With vs Without high expected medical costs given family size Small families less impacted With vs Without high expected medical costs after pregnancy Those expecting children more impacted Empirical Implementation Probit specification Prj0bchange 13 0 A H z HighRiSk Q HI HighRisk 27 with the accompanying test for job lock being 3 gt0 Data 1987 NMES Observed at 2 points in time 715 months apart 0 Know if at same job from beginning to end 0 14000 Households Sample Married employed men ages 2055 2978 individuals 0 Change 1 if leave job voluntarily Whether employed or not Results 1 With vs Without alternate HI Table 3 D in D Those with EPHI 30 less likely to move if they didn t have any other HI than those with other HI 2 Large vs Small Families Similar Results 3 Pregnant vs not pregnant Higher impact for pregnant women Problems Identification assumption is that treated with EPHI and With high expected costs have a shock that makes them less likely to move Having EPHI good high paying job Having high expected costs having a large family or Not having other insurance spouse With good job Pregnant Wife So the results could be impacted With having a spouse With a good job has independent effects Other Studies have used alternate data compared to Madrian and similar approaches With varying conclusions Issue Job Lock vs Job push Job push refers to those workers who leave jobs Without HI Criticisms 1 Having a spouse working with HI and the propensity for having alternate forms of H1 are not exogenously assigned Labor supply of husband and wife is jointly determined Husbands with working wives and spousal HI may differ in other ways Other studies have used similar approach but adding more detail on other attributes results not very different 2 Variation in availability of Government mandated continuation of covera e COBRA Get to continue to have firm coverage once you leave firm for 18 months You have to pay the price group rate so not totally free Federal Law 1986 Gruber AER uses same estimation strategy Theorv Health Insurance and Labor Market Equilibrium We already discussed taxbenefit linkage and implications for impacts on wages and employment Summers AER 1989 Incidence of Mandated Benefits Summers AER 1989 Examines impact of government mandate on wages and employment Setup Government decides that universal access to this goodservice e g health insurance workers compensation is desirable What are options 0 government provision o employer mandate Question posed by Summers Are the efficiency reasons to prefer a mandate vs public provision Are there distributional consequences Conventional wisdom prior to this Both are like taxes so for efficiency reasons the only thing that matters is efficiency in providing the service For government provision revenue must be raised leading to a DWL from the tax 25 Suppose Employer required to provide some bene t statutory incidence on employer taX or mandated insurance W W l W 2 D D 1 13 0 Before mandate L0 W0 Introduce mandate Shift left in demand D by the cost of bene t C Result lower wage and employment level 7gt W1 Ll Summers insight IF workers value this benefit then a job at a given wage becomes more desirable This leads to a shift out in supply by the worker valuation of the benefit orC where or is the valuation by the worker Result Further reduction in wages some offset of the decrease in L Case 1 Employee values benefit at cost 0tl W falls to bear the full cost No change in employment Full cost shifting Case 2 Employee values benefit at less than cost 0tltl Wages fall but by less than the benefit employment falls Lower DWL compared to pure taX TaxBenefit Linkage Key is if they value the benefit then not a pure taX And the shifting out of S can reduce the decrease in L If workers value the insurance at greater than cost ie they are risk averse L may actually increase 27 Why is there no taXbenefit linkage for government provision people paying taxes and people getting benefits are different no linkage harder for government to tailor plans to meet diverse preferences Overall predictions of employer mandate Increase in benefits lead to an unambiguous decrease in W While the impact on L is ambiguous TaxBenefit Linkage Key is if they value the benefit then not a pure taX And the shifting out of S can reduce the decrease in L 28 The Incidence of Mandated Maternity Benefits Gruber AER 1994 Background Mandating benefits provided by employer Efficiency Argument I Government provision requires raising tax revenue and such taxes create a distortionary efficiency loss I With employer mandates wages may adjust to the costs of the mandate but if employees value the benefit the efficiency loss is somewhat mitigated Note I Key to reducing DWL from the mandate is wage adjustment Wage adjustment requires 1 full valuation of benefits and 2 no wage rigidities I In Gruber s case we may have wage rigidities because the mandate is group specific women and antidiscrimination laws may limit wage adjustment I Illustrates more general point that mandates that are group specific may not have efficiency gains as advanced in Summers I Also for women with wages near minimum there is no scope for adjustment Contribution of a er lst em irical investi ation of Summer s hypothesis p p p g Policy Changes I Prel975 0 Coverage for pregnancy was not universal either not covered or limited 0 5075 of women had pregnancy benefits that were less comprehensive than benefits for other conditions I l 97 5 l 979 o 23 states passed laws outlawing treating pregnancy differently from other insured diseasesconditions I October 1978 Pregnancy Discrimination Act 0 Prohibited differential treatment of pregnancy in employer HI plans Goal Measure the effects of mandated benefits on wages for the group Outcomes Wages hours worked Labor force participation Advantages Easily identified beneficiaries of law women of childbearing age and their husbands Potentially large benefit 25 of weekly earnings Research Design 1 Analyze 19751978 period With state law changes Compare states With and Without mandated maternity benefit laws Difference in Difference Difference in Difference in Difference Affected vs unaffected workers What does empirical model look like then 2 Reduced form model using predicted costs of benefits Use second data set to predict expected mandate cost as function of Age specific cost of maternity coverage Probability they have insurance Type of insurance 3 Analyze 1978 Federal Law Caveat Federal law more comprehensive Treated States States Without law yet Control States States already having law DD and DDD as in 1 Like abortion literature experiment and reverse experiment Estimating costs of providing benefits Necessary to measure Whether full adjustment obtained premium calculator from insurance company input demographic characteristics can observe impact of adding maternity coverage TABLE l THE GUST m ADDING MATERNITY Emmi 0 TD A HEALTH INSURANCE PACKAGE Range Of 1395 A Of wages Cost as percentage nf Annual cost Ann uall cast 1975 weekly Cmrerage Demugraphic gmup 19 dullars WE dollars earnings seems to be consistent with Family Z thgyearold females 5934 35M 45 Family 3039yearold Tamales 5156 27 35 500k 0fthe Individual 20 29 ycar uld females 5321 H19 15 Individual BU EQyEErDld emales 252 392 39 enveZOPe Family 20 29yearDld males 5934 360 29 061101116 10 OfCOSt 1 39 3 e H I 39 A Family 70 39 year old males STSB 7 277 1 Ofchlldblrth gtllt probability of giving birth in a given year Data CPS 1978 1979 Before Federal law 1981 1982 After Federal law 1974 1975 Before widespread state law adoption 1977 1978 After widespread state law adoption T reared Married women ages 2040 married men 2040 single women 2040 Control Men Women gt 40 single men 2040 Key Are these a good control group Do these demographic groups have similar trends Identification comes from differential trends by demographic group within states Look back at Table 1 to see premium costs for controls vs treatments Experimental States 0 3 of the 23 states that passed mandates IL NJ NY 0 Identifiable in CPS passage enough before federal law to see impacts passed in same time period Control States Results 1 Bene t implementation across states DD and DDD Pre federal mandate Table 3 Unconditional DDD 54 fall in wages for married women Table 4 Conditional DDD educ exp sex marital stat nonwhite union industry occupation TABLE 3 DDD Earmares OF THE IMPACT OF STATE MANDATES mar HOURLY WAGES Locationyear change Change A Treatment Indiuiduals Marn ed Women 20 4I9 Years 0M Hcfurclaw Afterlaw Time difference 39fnrlocatiml Experimental slates 1547 1513 34 1012 0012 0017 1400 1496 Nomexpcrimental states 1369 13517I b1323 mow 173010 01314 14810 16443 Incatinn difference at a point in lime 11173 H In 0016 0015 Di erenccimdi mence D 2 H022 B Control Group Over 4 and Single Male 20 40 Experimental states 1759 1748 101 MINI 1007 mom l5624 5437 Wuncxperimenta states 1631 1627 L003 0007 0007 0010 4959 4928 Lucatinn difference at a point in time 1129 0121 0010 0010 Di ercnceindl erenm 0008 H014 DUI 41054 00215 Treatment effect is 54 fall in relative wages of 2040 yr old women Seems large relative to results in Table 1 and taking into account that not all women have insurance DDD across treatment groups Tame 4 TREATMEHTDUMMY RESULTS ACROSS DEMGGRAPHIC GROUPS P t What does this imply crmn age Lug Log Employment changes in 0 valuation says Group V lmurlar wage haursfweek prorwa labor input W a g e S Sh oul d Married women ages 20 41 1043 004quot a 39 140 39 H023 0022 MEII COSt Wlth n0 MUM accompanying Single wornen ages 20 40 LIME l104 01395 595 h 1 b 0026 0024 mm C ange In 3 0T imam Married man 3325 10 40 9 FIRE 0030 11139 9 LOB supply 11013 110153 o Compos lon of MI trealmcnts a 11323 002 0019 083 labor input may 1013 0014 NIB 39 m4 change fixed costs of work argument 9 PT i FT T I H r1 I Elli 932139 1335 uE 34 TREAT 3M x r 36w x TREATJ ma x meat 4 335 4 r x TnEmrf y 2 Reduced form model with predicted bene ts Predicted Cost expressed per week PrEmpl based insurance coveragePrFamily CoverageAgespecific cost of law change Data May CPS benefits supplement WCES medical care survey Private insurance data I am not really clear on the specification of the empirical model controls etc I think that it replaces TREAT in original model With the predicted costs Poorly written up Again a l in the wage equation implies full cost shifting Advantages individual rather than group level variation Disadvantages parametric valid model Tm 57 AND Luna mm REsULTs PARAMWWED 3051 or we MAW Sven cahon u 1i Am m v bog me Log wage Log Employment r uef ciam Log mg mu hunts minim hamsweek vmbn a 7 2 MI 7 mm 7 a 037 0004 7 1027 0759 mom in mm 00011 man 7 A1022 Sblfling Percentage an in 156 N u m 367 35868 mm was Rem11x 721 210 cost shifting Hours Worked increases Probability of being employed falls Problem When costs are not normalized by hours worked the cost shifting falls This implies that the wages of low hours workers were responding moreiwhich is not likely since they have low rates of health insurance 3 DDD using Federal Experiment 1m 7erme 125m TS PFDFRAL Emmst Speculum i u Aw v vi Dzmngrapmc Lag wage lug age lag Emplnymzm Change m min gmupcummem lag wage no noun minim hamsweek lprobil lahm mpm Mmmx mama ngzimrw fainzi 00012 mm 700071 10012 00098 0028 7 o 0055 Smxle women ages zwa 7 mm 0 0157 um c0219 H014 mum 0mm nms Manled men ngsmrw mm 70mm nn rams 00012 0 mm mm n 0005 All Hanna 7 0mm time n 0001 a 0033 0mm 00m 00213 mom Indwldual parameterilsuon 79537 70023 70017 700001 0007 700005 mm 001m mom 0 ms 0 was mums smnmmmemm 5 9a 75 Evidence of cost shi ing to Wages but at levels only at 50 of earlier levels Caveat federal law more expansive control states are partially treated 9 expect smaller impacts Overall 100 cost shi ing onto lower wages Much smaller declines in labor input consistent With cost shifting model Comments DD methods now easily handle treatments happening across states at different time eg welfare reform No need to limit to 3 states Seems like they hand picked the control states which is a little suspect Need to present graphs that illustrate DD findings limit to FT workers since PT workers often do not have insurance Hilary Hoynes UC Davis EC23O Taxes and the High Income Population New Tax Responsiveness Literature Started by Feldstein JPE The Effect of MTR on Taxable Income A Panel Study of 1986 TRA Hugely important paper 0 We care about estimating the LS elasticity because it is centrally important for optimal tax min DWL But our typical LS analysis is narrow employment hours If employmenthours do not adjust then should we conclude taxes have no DWL NO 0 This misses other important dimensions that may also be responsive to taxes 0 Intensity of work career choices 0 Form and timing of compensation e g iMTR gives incentives to move into taxed compensation such as wages relative to untaxed compensation in fringe benefits 0 Tax avoidance tax evasion Auerbach amp Slemrod JEL give the following hierarchy of tax responses 1 timing 2 financial accounting 3 real activities Key distinction income shifting versus creation of new income How do we respond to Feldstein s point Option 1 Estimate separate tax responsiveness on all of these margins Option 2 Estimate quottotalquot or quotnetquot effect of taxes on some broad measure of economic output Recognize that taxation does not affect the relative price of each of the quotgoodsquot so they can be treated as a Hicksian composite commodity and we can just look at the change in consumption or taxable income in response to tax rates to gauge the magnitude of distortion Feldstein quotTax Avoidance and the Deadweight Loss of the Income Tax RESTAT 1999 shows that under certain conditions the overall elasticity of taxable income with respect to the net of tax rate lMTR is the relevant parameter for examining the impact of taxes on revenue and welfare This is how the literature has proceeded 4 Difficulties in the empirical analysis of these issues 1 Choice of control group If we think that trends or elasticities are higher in the high income groups then it can be hard to find a valid control group Not using a control group is not desirable either Elasticities will be biased upward if for nontax related reasons top incomes increased more rapidly than average incomes during that period gtPossibly contaminating factors skill biased technical change increase in demand for high skill workers international trade decline of unions general increase in upper tail inequality 1 Comparison of years just before and after the reform may capture a short term response when we want the long run elasticity for welfare calculations 2 Analysis of high income folks presents particular data problemshard to find large samples of rich folks topcoding and hard to find data detailed enough to characterize various components to income Simple model Maximize uc 2 subject to c 21 T R Increasing in c after taX income consumption Decreasing in z before taX income economic activity R nontaxable income Leads to quotreported income function Zl 239 R If income effects are small then the reported income function only depends on the net of taX marginal taX rate lT Key z is not just labor supply but a more comprehensive income measure 82 1 239 81 2 Z elasticity of interest 6 FACTS HIGH INCONIES and MTR Top Marginal tax rate Year Top MTR 1960 91 Mid 19605 70 ERTA 1981 50 TRA86 28 OBRA90 31 OBRA93 39 6 2003 35 Broad trends in income Saez TPE a top 1 tax mm A Baumn wx a mm mumrum g s s e 3 mm m Rm Mammal m m mame shlm Feldstein 1995 Innovation was using panel data His thinking was that this would solve the problem of the T group changing composition over time But since have learned that panel data can create a mean reversion problem Those super high income in one period are less likely to be high income in the next period Examines TRA86 Data Panel of tax returns taxpayers in 1985 before and 1988 after Nonaged taxpayers married in both 1985 and 1988 Model relate changes in taxable income to changes in netoftax rate LE 1 Rurast 0F Tuuu Ixcom or Nomcm Mum TAxmvzxs 1391 Emma m Mmcmn TAX Runs 142mm 1955 AND 191111 Pancmn z Camus or Adjusted AG M1111st Adjusmd Taxable 1915 AG Nut 11 Adjuned Exclullmg Taxabl hmamc 21m 1335 1mm 501101 ObaleATmss Tax 1131 run AG Capilal Gains 11mm Gross L055 TAX RATE 1 2D 5 4 l5 t 7 22 3117 300 90 94 114 13 a 134 25 5151 9119 133 45 24 35 37 211 427 7151 163 117 50 5111 33 515 771 87 22 25 25 as 675 515 151 111 9 5 153 42 9151 152 211 117 2211 225 15 1269 45 3011 1411 155 151 49 1777 35 412 2911 427 3119 50 47911 22 44 11 71111 924 511 22411 55513 122 46 62 64 42 45 197 255 147 210 2113 19 a 422 557 716 1411 munquot General pattern The larger the increase in taxable income the net of tax rate the larger the increase in Net of tax increase means marginal tax rate decrease Elasticities using DD analysis relying on variation in size of taX reduction across groups as in Eissa s married women LFP paper T ESTIMATED ELASTICITIES or TAXABLE INCOME WITH RESPECT To NETOFTAX RATES Taxpayer Groups Classi ed by 1985 Marginal Rate Adjusted Adjusted Taxable Net of Taxable Income Plus Tax Rate Income Gross Loss 1 2 3 Percentage Changes 1985 88 1 Medium 22 38 122 62 64 2 Hi h 42 45 256 210 203 3 Highest 49 50 422 716 448 Differences of Differences 4 High minus medium 134 148 139 5 Highest minus high 166 506 245 6 Highest minus medium 300 654 384 Implied Elasticity Estimates 7 High minus medium 110 104 8 Highest minus high 305 148 9 Highest minus medium 214 125 Estimated elasticity is HUGE ranging from 13 Critique of Feldstein Use of panel data keeps quothigh income sample static 9 no dynamic sample selection like we talked about With Eissa TRA paper However it is still possible that there is a differential trend for this static sample of high income folks Without a valid control group the estimates are still biased If you want to capture the dynamic response that upper middle income folks change behavior and become high income in response to the policy then Feldstein s approach Will miss this Classifying 39treatment39 group by pre income mean reversion Simple DD model masks SR and LR responses to taX Saez quotReported Incomes and Marginal Tax Ratesquot Tax Policy and the Economy 2004 Examine effect of MTR on reported income by looking at the share and composition of income accruing to the top of the distribution Examines tax responsiveness Over time 1960present Across the income distribution Using public use tax files 19602000 and TAXSIM Hugely innovative use of tax data to study these problems original pub is Piketty and Saez QJE 2003 Measuregr0ss income rather than taxable income as in Feldstein before adjustments and deductions excludes realized capital gains governed by different MTR Outcome variableslevel and shares of income by type accruing to top income groups TABLE 2 Thresholds and average incomes in tap income groups in 200i Percentile Income Income Number of verage income threshold threshold gmups tax wills in each group ll 3 33 Ht 5 Full population 13353911100 42309 Median 251176 Bottom 90 120230100 526615 Top 19 5813 Top 1059 69450 100489 Top 5 120212 Top 51 39equotu 5343560 16236b Top 11 2933 Tup 1 05 551945 33279170 Top 5 397949 Top 05 01 5356 E li iil Top 1 3 3 1134349 Top Dl 1 1213230 EDEHBOI Top Dl f f 534995 Top 001 13359 13015242 Components salariesoptionsbonuses Scorp income Schedule C partnership dividends interest other TAXSIM individual income taxes only no FICA corporate income tax Methodology Time series regression of logaverage income or logincome share on lognet of taX rates and trends Do this for different income groups to capture heterogeneity TABLE 3 Elasticities af income with respect to netaftax rates in the aggregate bottom 99 and tap 1 Regression Eegmassiun Results Regressian in levels in levels in levels time Central time controls elast1c1t1es 1 L 31quot concentrated at Panel A all tax unit3 1 TI 1 HD the top of the El39 139 39l 4144 ll L4 as m y 034 43331 055 d15tr1but10n Time trend I39es 25 Time trend square as Panel B lmizom 99 m mzii s an 4 43 M sensmve Elasticity 085 l l to Inc n 39 07m in was us 0 Of Time trench its Yea tune trends Time trend square quotes Panel C top 1 tax units 39 Elasticity 183 071 059 037 1322 111le Time trend Yes quot1 95 Time trend square Yes A A 1A uul l39FJ1l39HI IIII UIDI391 ELEEJEEIE EH1 Model is specified as a function of NET OF TAX rates lMTR so we expect the elasticity to be POSITIVE Other results in paper Little impact even in the top decile Small changes to 1960 reforms Large changes to 1980 reforms Key change in compensation in 1980s away from corporate sector and towards individual taxes partnerships subchapter S corporations Key attention of short run vs long run responsiveness SR response of wage income to 1986 and 1993 When he does IV he uses the top MTR as an instrument for the average MTR TABLE 7 Elnsticities ofillcame slim 39 39 r um WK quot2 gm Newcy ewey Newey Newey est 015 West 015 ZSL39S est 0L5 West mgrcssiun ion regression Regression iegrcssion rugnssion no ti with hme with time a time with lime wtth time cuntml nu controls urinals conlmls cont1075 W 3 4 5 A Tap irime gmups 8 Intermediate quotICU71139 Sump Tap 10 Top 104 Elasticity 077 033 032 4 011 004 036 008 005 017 009 010 Firststage t statistic 994 65 at instrument Top 5397 Top 5 1 Elasticity 125 043 039 014 012 009 030 009 005 0 23 004 001 Firststage tstatistic 105 816 ui instrument Tap 1 Top 1 5quot Elnatiuuy 150 062 059 092 030 029 028 012 008 021 0024 0 07 FiKSstuge tstalisiic 1011 065 or instrument Top 05 Tap D5 01 Elasticity 155 072 059 121 052 049 05 013 009 022 009 003 rim stage mutian 9 i 7 7 of instrument 93921 Elasticin In 01 quot1 q Top 01 001 0 1 9 0721 075 027 019 111 011 a 5 n m Firststage psttttislic u 37 969 n inslrumcnt Elasl TOP 1 Tap 001 145 103 109 145 1 08 109 036 0 32 0 6 036 032 016 Firststage tgtlanstic tam 1801 of instrument oggz oga 90 m vulgww pig 53100101 Wm TABLE m Elasrin39h39zs of mg mmm 51mm and 725ch m llclnltax mm m various upyquot wag income gravy NeweyM e NeweWcsI NwwLyWcst gt12 me 015 015 015 05 mgmmm ngessmn regleision rcgmssmn no am with mm a time Mh um mum s Lumml mums carlmls I l A M mug Inmm was a lnlmvmlmr gmups Twp M Top no5 Elmslxmy 010 n m 4143 4w 0 5 m m mm m 31 Top 5 Tap 39u Elaslkiry 041 0J7 417 007 1155 um mm m 02 Top 1 Tap 15 Elashcuy m7 0 39 o 31 n 5 may m m cm 005 Tap 057 Tap ls01 Emmy 2 33 L50 Ms 1054 032 005 Tap 11 Top 0100 Elasmny 251 2 15 u 71 4o u 7 n 11 Twp now Top mow Ehmkily me a 09a 10 050 m 42 Nutn Lulqu mwm mums murmuu a In an mg hum mm m 3 Mann m w Big caveat in this work Ultimately this is a sort of quotsingle difference model looking at the time series of taX rates and the time series of income shares There are other factors that can affect these outcomes that are not being controlled for here Compelling data but not a conVincing research design As demonstrated by the sensitivity of results to including a simple time trend Note that this suggests that elasticities are concentrated at the top of the income distribution but this does not imply that elasticities of labor supply or earnings are heterogeneous or concentrated at the top of the income distribution 9 This is an area where more work needs to be done Tie this back to labor supply analysis or basic earnings analysis to see if elasticity varies across the distribution in this way 20 A Goolsbee What Happens When You Tax the Rich Evidence from Executive Compensation JPE 2000 example of new tax responsiveness literature also fits into large literature on behavior of high income persons taxes Deals with some of the criticisms of the original Feldstein JPE 1995 paper Goal of paper Estimate elasticity of taxable income to MTR Separate temporary permanent responses to MTR changes 21 Tax Law Change OBRA93 T MTR on high income Think back to the criticism of Eissa s paper If income is trending upward for high income folks and near high income are not a good control group then this research design will lead to a downward bias in the effect TRA86 Eissa MTR i for high income folks inc trending T 9 Bias OBRA93 MTR T for high inc folks inc trending T 9 Bias Tax changes 1 marriedjoint 31 to 396 for 2 250K 31 to 36 for 140 250K 2 Eliminate cap on Medicare payroll tax 2 130K 3 Eliminate deduct of corporate exec payments 2 1 mill unless performance based 4 Anticipated since Clinton ran on tax increases So can we capture anticipatorytiming effects 22 Data Publicly available panel data on executive compensation 19911995 Top 5 executives in corporation in SP 500 companies plus other mid small cap companies Sample selection sample includes only individuals W 2 4 years of data m balanced panel Survivorship bias Advantages of data a huge samples on rich 21299 person years b Income components Salary bonus options exercised long term incentive plan multi year bonuses other c good to look at timing of response as some options are easier than others to adjust Disadvantages of data a have to impute taxable income do not have full household income capital gains spousal income measurement error 23 General points important for methodology Constructing treatment group prepost Want to look 1 2 years prior to taX reform especially in this case when tax hike was anticipated Some of these income sources are highly variable Mean reyersion if report lots of options in t 2 e g in treatment group high income in pre period Then back to normal level int 9 observed income obseryed taX rates Spurious correlation Solution Use 5 year average to get permanent income to construct TC groups I think this point is only valid for using longitudinalpanel data If pooled cross section then someone will always be in high group 24 Methods Anticipatory model I lnincomelt 2 at 8 lnl taxml 5 lnl taxit X it R git How do deal With zeros in some income categories How does this deal W endogenous tax rate he does not talk about it individual fixed effects longitudinal data oci controls for fixed unobservable preferences tax reform As 9 taxit variable Expectations B lt 0 anticipatory future T taxes ll tax 9 move tax income to today T 5 gt 0 SR current T tax 9 behavioral response less income SR elasticity Total LR effect B 5 lt 5 SR elas B 5 in elasticity form log log specification 25 TABLE 3 mmmsu or TAXmm INchr Fum nmusc Na No Yes 11 141 51 mu 7 1ax 1st 1173 1427 126 3241 536 mu mm 71555 555 1n11 uxx 1gt 1 15m 187 In marktl Iluc 1194 1117 awningsamt ma 1 1251 Time 1er 0071 Taphmkkl x lune 1155 1103 111101 11115 Tapbmnkcl x Hunkname 174 019 Topbmckul x unnungs 2112 1401 ch dummies no no rm nD ya ya bserva onx 15395 15477 151135 11493 211407 144251 1quot 73 77 77 117 112 117 Results Table 3 l 3 all treated executives average inc 2 275K SR response 5 12 spike in SR Anticipation effect 3 08 total effect E 4 small in long run 4 6 include all execs highest lower income treated are those at high inc level year dummies included identification variation across people in sample cross sectional Treated X Time Not typical specification Robust results 27 Table 4 Stratify by income group most responsive on very high end those woptions no effect when options taken out last column Table 5 Components no logs possible since Os Levels option s are most of action Conclusion Smaller behavioral response than older lit Critigue l are these people representative of overall high inc upper bound 2 Survivorship bias 3 same concerns about validity of control group 28 Hilary Hoynes UC Davis EC230 Redistributing Income Through the Tax System EITC and Labor Supply Outline of Lecture 1 Overview trends and program details 2 Economics Theory of labor supply and eitc 3 Empirical studies Eissa and Liebman 1996 Meyer and Rosenbaum 2001 Rothstein 2007 Eissa and Hoynes 2004 Introduction EITC provides cash transfer to low income families With children Transfer provided in the form of a taX credit Refundable taX credit EITC has been in place since 1975 originally intended to offset cost of payroll taxes for low income families Through tax acts of 1986 1990 1993 it has expanded into the largest cash transfer program for poor families Because of its size it is important to understand how the program impacts the poor 1990s Period of tremendous change in terms of government assistance and low income families 0 Reduction in support through quottraditional welfarequot AFDC 0 Increase in support through quotin workquot bene ts taX credits EITC Figure 11 Real Spending on the EITC Billions of 2003 Dollars 50 fI rquot 40 fm 30 8 5 20 Av 10 I AFDCTAN F Expenditure i IN EITC Expenditure 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1975 1980 1985 1990 1995 2000 Costs of EITC compared to other programs for the poor EITC TANF FS Cost billions 33 4 245 210 Families millions 196 21 74 Average Bene t 142mo 3 51mo 174mo 2005 Annual Employment Rates for Women By Marital Status and Presence of Children 19832006 95 i E M gt1 0 9047 vzquot E a 85 7 a N a 80 7 2 Q 0 E7547 a 5 0 z g a 70 7 Single No Children quot 9 MarriedNo Children 650 i S1ngle Ch1ldren Married Children l l l l l l l l l l l l l l l l l l l l l l 1983 1986 1989 1992 1995 1998 2001 2004 Source Eissa and Hoynes Tax Policy and the Economy 2006 EITC Eligibility at least 1 child lt 18 in the family available to both single parents and twoparent families positive earned income earned income and adjustable gross income AG1 below maximum amount EITC Bene ts Credit Amount phasein or subsidy region Credit 73 E 0ltEltEO at region Credit C max EOltEltEl phaseout region C Cmax 239P E El ElltEltEmax TS subsidy or phasein rate varies by kids IF phaseout rate varies by kids Cmax Maximum credit varies by kids Marginal Tax Rate Marginal tax rates landscape and EITC 1 child no amt 2004 50 40 30 20 10 0 10 20 30 40 50 25000 50000 75000 100000 125000 Earnings Single 1child Single 1child with FICA 150000 After Tax and Transfer Income Phaseout Region Slope wIrp Flat Region Slope w Phasein Region Slope w1rs Expected Impact of EITC on Labor Supplv Labor Force Participation Hours Worked PhaseIn Flat Phaseout Above Phaseout 5000 4500 4000 3500 3000 2500 credit amount 69 2000 1500 1000 500 0 4716 Married Couple Schedules 39339quot 2853 n n I n a n n I n a n n I i l39 40 428 35241 R19 0 5000 10000 15000 20000 25000 30000 35000 40000 earned income Two or More Children One Child No Children EITC Credit 1996 4000 3000 2000 1000 Illustrating differences in expansions by family size 1996 dollars Families with 1 child A Schedule for Family with 1 Child Families with 2children 1996 EITC 1993 EITC 1990 EITC x 39 39 1984ETC V 39 0 a 0 5000 10000 15000 20000 25000 Earnings 1996 s 30000 EITC Credit 1996 4000 3500 3000 2500 2000 1500 1000 500 B Schedule for Family with 2 Children 1996 EITC 1993 EITC I 39 1990 EITC 0 39 39 1984 EITC 39 Q 1 0 5000 10000 15000 20000 25000 Earnings 1996 s 30000 Issues to think about in tradeoffs between AFDC and EITC Ef ciency labor supply distortions Family structure distortions Administrative costs Outreach Equity issues Who is getting aid Stigmatakeup The EITC and Labor Supply Main Empirical Studies Single parents Eissa and Liebman 1996 Meyer and Rosenbaum 2001 Rothstein 2007 Married Couples Eissa and Hoynes 2004 Challenges to estimating the impact of EITC on labor supply national program no Within state variation inside tax system takeup issue are families Without children a valid control group Variation to take advantage of tax law changes created discrete changes in credit in certain years tax law changes differentially impacted one vs two child families EITC expansions impacted differentially across earnings groups Literature mostly focuses on impact of EITC on the labor supply of single women With children Eissa and Liebman OJE 1996 Early differenceindifference paper rst paper to examine behavioral impact of the EITC Examine 1987 expansion of the EITC phasein rate 11 gt 14 maximum credit 550 gt 851 Quasiexperimental approach Compare labor market outcomes hours employment of those affected single mothers to those unaffected single women without children Difference is EITC Data CPS 19851987 before 19891990 after Estimation y 05 X yOELIGZ ylPOST86t yZELIGZ POST86t git y outcome variable labor force participation hours worked Z demographic variables ELIG1 if woman has child 0 otherwise POST861 if yeargt1986 0 otherwise The full speci cation combines periods and compares changes in outcome of treated with kids to controls wo kids X added to control for any observable differences between the groups yl controls for shocks that affect both groups 70 controls for the permanent differences between groups 72 is the treatment effect Expected sign Identifying Assumptions 1 no contemporaneous shocks to treatment and control groups over the period 2 no underlying trends in two groups over the period Treatment groups Single women with children in lt12 or lt12 years of education Single women with children predicted to be elig for EITC Control groups Single women wo children with low education or low pred earnings Single women with children and high education or high pred earnings Results Unconditional DD Table II Some attention to pretreatment trends see Figure 11 Need to adjust scale to examine this better Main results Table III Participation increased by 19 to 28 percentage points Sensitivity checks nice idea Figure 111 Estimate model with full set of year dummies and interactions of year dummies with treatment group dummy Fig 111 plots the childyear dummies Show tuming point in 1986 Reverse trend Could it be another factor affecting women with children Labor market AFDC Little impact of adding this to model No effect beyond in gthigh school sample or with predicted income above phase out Table IV Hours worked estimate conditional hours equation dynamically selected sample At what level of hours do new entrants have Less robust impact on hours or positive Table V TABLE II Sample ell unmarried women Without Demographic Unemployment Stnte Second child Separate year eevnriates characteristics end Ame dummies dummy interactions Van39ables 1 2 a 4 5 0 Coef cient estimates Other income 1000s 7 70035 001 70034 001 70034 001 0034 001 0030 001 Number of preschool eliildren 70305 010 70 270 013 7 201 018 278 010 70 010 Nenwhite 7 70 422 010 70 521 030 510 031 Age 70 237 059 7o 200 060 70193 060 Age squared 7 0 007 002 000 002 0 000 002 Education 70 020 014 70020 014 7o 020 014 Education squared 7 0010 001 0010 001 o 010 001 Secnnd ehlld 0113 040 70117 040 State Unemployment rate 7 70090 007 70063 012 70054 70064 12 State Unemployment rate kide x i e 7 0028 010 0029 010 0029 010 0030 010 Maximum 11mmth AFDC bene t 7 7 70001 000 700010000 70001 001 70001 000 K 71053 020 70250 029 71430 100 71450 110 7 1402 110 P0 0 700010201 0019 031 152 007 0104009 70094 000 Kids x 1705150 7 0000 027 0074 030 0103 0371 0113 0371 0007 0431 x 1980 0030 057 s x 1900 0116 050 x 1990 0112 057 Second child gtlt post 0051 043 Leg likelihood 720750 717105 710703 710633 716520 000 029 Predicuzd participation response for treatment group 019 000 026 010 023 000 022 009 015 015 Less Than High School 1992 Doilars Participation Rate Deviations 14 00 39200 Maximum EITC 0 1000 800 L 005 04 4 00 Marginal Effecis 200 015 o I I I I I I l IQBi 1985 1985 1987 1989 1991 FIG Maximum EITC and Marginal Effects from IUD X YEAR Dummies Meyer and Rosenbaum Welfare the EITC and the Employment of Single Mothers QJE 2001 Contributions of the paper Typically studies consider the impact of one program on labor supply This study models the impacts of a comprehensive array of programs They introduce a way to combine the nancial incentives of many different programs that change the returns to work They do this without assuming that wages are exogenous The study focuses on a very important period that has seen remarkable gains in employment of single women with children AND is a period of massive changes in public assistance programs Presented as a psuedostructural model with focus on using exogenous variation policy changes to identify key parameters The facts annual labor force participation rates of single women with children rose over 9 percentage points between 19841996 They rose even more for low educated single women with children These gains were not present for other similar groups See Meyer and Rosenbaum National Tax Journal 2000 where the authors conduct a difference indifference analysis and consider a whole host of control groups Economic model of work decision Prw0rk PI UYWLWPWX gt UananinX U utility function Y income in work w and nowork nw states L leisure in both states P participation in welfare capturing cost of participation X demographics The nonworking choice has no uncertainty income and leisure are known takeup of all bene ts is assumed to be 100 The working state has uncertainty about wage rate and hours worked Wages and hours are drawn from the empirical distribution of workers no selection No information about differences in wage opportunities by education etc is specified The 100 takeup rate is maintained for workers all eligible workers participate Keep in mind this model and all of the estimation is applied to a sample of women with and without children Think about similarities to a DD model Functional form Assume utility linear With additive normal error term UYLPX aY LpPyX8 Taking the difference in utility the probability ofwork is Prw0rk PraEYw ELw pEPw wa 8w gt 0 an anw 7an 8W PriaEYw m ELw LW pEPw PM 7X gt a CD06EYW an ELwl an pEPwl PW 7X Issues Structural Utility function implies leisure and income are perfect substitutes Uncertainty in work choice Uncertainty in wages and hours worked They assume all persons face the same bivariate wagehours distribution Then they take expectations of Uwork with respect to some empirical distribution of wageshours Mechanically they take a sample of women with annual earnings of gt500 and 10 hourly wage groups x 5 annual hours groups Create expected earnings using Where w and h are midpoints in each range P is the percent of persons in that wagehour cellempirical distribution Note that this expectation is identical for all women in all years Does not rely on ANY demographics any crosssectional variation for identi cation of program effects Issues selection bias Can new workers get the same wagehours distribution as workers Ultimate estimating model Allow for differing coef cients on different forms of income stigma ii different coefficients on income if working or not working Given linear U taking expectations and same distribution of wageshours 100 takeup gt earnings leisure and participation if no work is absorbed into the constant term Assume women with children all participate in welfare if eligible women wo children never do This generates the following model they estimate Prw0rk CD01Etaxesw 052EAFDCFSW 053EMedicaidw pEPw 04AFDCFSW aSMedl39caidnw 7X Expected signs of coefficients Positive 12 0L3 0L4 15 Negative 11 p Note that each of the tax and benefit variables are averages across the wagehours possibilities They also vary by state number of children etc Tax Transfer programs modeled in budget constraint Federal taX liability EITC other income Tax acts in 90 93 State tax liability EITC other income AFDC amp Food Stamp bene ts Welfare reform variables state waivers Medicaid expansions TrainingEducation in AFDC program recipient TrainingJ ob Search om AFDC Program recipient Child care assistance on AFDC Data Two CPS data sets 19841996 March Annual Demographic Survey provides data on the previous calendar year any work weeks hoursweek earnings income Merged Outgoing Rotation Groups MORG provides data on the previous week any work hours earnings VariationIdenti cation Variation in key RHS variables come through policy variation State number and ages of children year Results A Unconditional DifferenceinDifference Table II Compare single women With and Without children Can not identify the source of the changes in employment Employment of those With children increased relative to those Without children What about stratifying by education level B Conditional DifferenceinDifference Table III Add controls for race age education number of children uneamed income uratechild state Main effects year year anychild Increases for single women With children relative to those Without children especially since 1991 C Structural model Table IV Same controls as conditional DD add policy variables if all income is the same in utility then coef should be equal Given that takeup is lt100 especially for workers this amounts to scaling up the coef on welfare benefits if work and stigma if work This is consistent with the results Taxes and welfare have large marginal effects Medicaid little Smaller effects when limited to women with children How different is this from a differenceindifference larger effects for lower education groups Contributions of Policies Table VI Decompose observed 19841996 and 19921996 8496 increase 62 EITC amp Taxes 25 AFDCBen 15 Waivers 9296 increase 27 EITC amp Taxes 17 AFDCBen 15 waivers ORG voxked last Week 1 Yam 0 Education All lt 12 12 gt 12 Explnnmm39y variable 1 2 a 4 Income tzn39es ifwark 70027 700417 7 01s in 0 s v 00034 00030 00051 1 are 11 111mm 02090 7 031 700495 2 005 139 00024 00030 00030 Welrm bene t IfWDl 1 0772 00054 0539 in moonsyea 00073 00171 00109 Pmbability of AFDC receipt 701985 7 29 0 70 1027 if work 00230 05 2 00303 Medicaid if Work 7 0009 0040 7 0 0107 in Momsyear 00033 00000 00055 39 it 00130 00405 7000 00071 00100 00102 1 37 t nations 0222 0355 001 Indicator variable 00110 00260 00158 Trainingi 39 7 030 700524 00563 in 10005lyuar 00100 00010 00283 Trainingijob searchother 04 0047 060 026 in 10005year 00117 00272 00190 00175 Child care 00227 0027 00190 00226 in 10005year 00065 00142 00104 00104 Stale unemploynmm me 700100 700092 700101 700105 in Jeni Luge puinla 000071 00020 L00013 00010 Any children i smte unemp rate 7000 1 00009 700010 0002 in 99139chng points 00009 00021 00014 013 Number ofobsemcions 373002 51146 134432 100054 Table IV TABLE VI CONTRIBUTION OF I LA 1mm 0 mu RELATIVE 1 1 m m MOTHERS VERSUS SINGLE WOMEN 00711007 mLDKEN 198471996 AND 199271996 190471990 199271990 ORG Mmh cps ORG March cps Explsnawry 00110010 A in Emp oz of mm A m emp 0 000001 A 1 amp 00 00 00001 A 111 amp 0 of 000 Income mes 0 Mark 00430 00720 014 00100 00005 0510 Welfam mumnm 000000 00179 00150 00114 00099 1140 Welfam bene t 0 work 00005 00003 0 0045 700000 7 30 10000010 00 AFDC receipt 1f work 7 00002 70 000 00020 00017 20 0 001cm 39 war 0 0070 0 0002 00023 700011 7120 1000 welfam bene ts amp Meuhcmd 00112 00125 00000 00070 510 quverwnytune Lunit 00054 00075 00052 00073 050 Wa er nytermmations 00040 0 0090 00010 00099 11402 0 welfare Wmvel39s 0 0099 00174 00093 0 0172 19 00 T7 ammggeduczm r 0 0101 70 0000 7 00021 70 0020 r 2 sq Traxmng 0 snurchmhcr 00005 00077 00047 0 0050 0 40 Chi care 00000 00009 00011 0 0010 1 5c 0001 trammg amp 00110 mm 00032 00050 00009 0 0000 09s Demugraphms 700073 70 0000 00107 0 0112 12 00 00000 00172 00190 00150 10202 Tm 00705 0 1 174 00091 00009 10000 Things to think about linear utility 100 takeup of welfare single wagehour distribution Similar to DD Why not include controls for kids year There is still identi cation in the trends across these two family types I suspect that this is a large part of the variation in the tax and transfer variables When they limit the analysis to mothers only Table V the results are smaller and in some cases lose significance Teaching the Tax Code Earnings Responses to an Experiment with EITC Claimants by Raj Chetty and Emmanuel Saez starting point of the paper is the observation that there seems to be more sensitivity on extensive margin than on the intensive margin this is certainly well illustrated in the eitc literature However the literature really DOES NOT address the issue that when a policy affects the extensive margin then a conditional analysis of the intensive margin could suffer from dynamic sample selection their hypothesis is that this is due to a lack of information lack of salience in tax They set up an experiment at HampR Block Tax preparers randomly treat clients by telling them about the EITC and where they are in the schedule They then follow them to the next year Raj and Emmanuel s take on prior beliefs h Perceptions of EITC Schedule For these prior beliefs 39 39 Famequot 5mm the predictions are clear 1Phasein hoursearnings should increase 2Flat no change 3Phaseout 1 hours earnings 5000 10000 15000 20000 25000 30000 35000 40000 Should decrease EITC AmountS 1000 2000 3000 4000 5000 Eammgs Ks And that is what they found I 39 FIGURE IV Yeszamingsni h39ihn nn 39Complyjng T1 1 r 39 m for a 1 Dependent 6000 Em Amount st 4000 Earnmgs Densuy 2000 0 5000 10000 15000 20000 25000 30000 35000 40000 PUSLTreatment Vear 2 Eammgs 5 E m Trealmenl Effects on EITC moms and Eammgn Dlsmhullnn Dependent vauawe A Em Arm 5 A EvTc Am 5 Middle Inc W5 M 05 In 6 Law me my H gh Inc M A Eamlngs s 0 Controls 01 Comm w Comm w Controls wz Commas m 2 lt32 m 5 way 7 1mm Samara 240 W 17 0100 037 4161 293 N30303 v477 use a 54 05 0 34 O 45 L33 451 1 531 121 0091 077 71 an D 54 035 ax Professmms 725 50 05 90 V 71 r 4 2 A 2109 46 10 0 07 1045 0 75 123 56 p 191 p341 278 2961 73 01 71 A71 VI 601 3220 111 215 45 17x 24725 090 c 52 u we 57 12040y u 7 v 0 m 14 531 2 21 230 in 55 213 2 05 while I think the paper is really cool I am not sure what it is testing Maybe this treatment can change behavior But that is not very interesting But we ultimately want to put some interpretation on the results They want us to think that this re ects moving toward more optimal behavior by changing beliefs about taxes But I am not so sure This prediction hinges on knowing that the prior beliefs are This is unknown Other cool things happening with the EITC Powerful rst stage increase in employment and expansion of EITC means income increases Can we use this to examine impact of income on children s outcomes other family well being measures helpful since income is endog and it is hard to nd instruments for these changes Explored in the literature birth weight test scores Dahl and Lochner Eissa and Hoynes 2004 Journal of Public Economics Examines impact of EITC on labor force participation of married couples Incentives different than for single women Economics Using the secondary earner model the predictions for the married men are the same as single persons employment increases and ambiguous effects on hours Predictions for secondary earners will in general depend on the earnings of their husband Consider the predictions for the woman as a function of the earnings of the primary earner Earnings of Effect on nonlabor Effect on rst hour Predictions for Prwork husband income of woman net of tax wage for and hours woman Nonworker none increase PW0rk Hours 7 E in phasein increase increase PW0rk Hours 7 E in at increase none PW0rk Hours E in phaseout increase decrease PW0rk Hours Qualitative prediction is that the EITC will lead to a reduction in employment and hours of secondary earners Data CPS 19841997 covers 3 expansions in EITC sample of married couples With educlt12 years ages 2554 2 Estimation Strategies 1 QuasiExperimental OBRA93 compare change in labor market outcome for treated married couples With kids to controls married couples Without kids 0 largest ever expansion in credit 0 created large relative increase for families with 2 or more children Identi cation 0 kids exogenous random treatment assignment 0 no interaction bW group and time other than credit expansion 2 Reduced Form We were concerned about whether married couples with children were on same trends as married couples without children So we present an alternative identi cation strategy using ONLY the sample of married couples with children Take advantage of additional variation time series expansion in EITC due to TRA86 OBRA90 OBRA93 crosssection effects should be greatest for those with larger families lower wages women married to low earning men Tax Simulation Routine EITC other federal taxes payroll taxes Assumptions secondary earner model holds gt income wage and tax calculations for the woman take into account the earnings of the man exogeneity of gross wages Estimate probit model for probability of work PrPlt 1 2 a 732 w t3 5Zit state time 8 y net of tax nonlabor income includes husband s earnings Wlt average net of tax wage also includes husband39s earnings average tax rate from 0 gt pt or ft status Findings qualitatively similar results for both methods small insigni cant positive effect on husband s employment modest and significant negative effect on Wife s employment Rothstein quotThe Unintended Consequences of Encouraging Work Is the EITC As Good as an NIT Evidence from Eissa amp Liebman Meyer and Rosenbaum and others consistently shows that the EITC has led to increases in employment among single mothers with children If so could these increases in employment lead to a reduction in wages If so this could reduce the bene ts of the EITC among those eligible and reduce wages for ineligible workers competing in the same markets as eligible workers EITC vs NIT NIT has guaranteed income transfer if no earnings EITC is only available to workers Besides evaluating the tax incidence of the EITC the paper provides semiparameteric estimates of impact of EITC across wage distribution AND contributes more generally to the empirical tax incidence literature Identi es through 1993 policy expansion largest to date Focus on women CPS MORG observations on hourly wages and hours worked last week Prereform 19921993 Postreform 995 8 97 TAXSIM and March CPS used to simulate ATR and MTR for groups MTR at current earnings ATR calculated comparing taxes at current earnings amp zero earnings calculate mean MTR and ATR by skill level wage level and demographics marital state X number of children 012 Use DFL reweighting method to make post reform sample look quotsimilarquot to pre sample F3 115 3 Change in mean ATR among famihes with Woxkjng VOIDED by skill and group a m mm quot Unmmned ldnkl U1mmmed2cln1dxen a h Mr Manned 1 chdd n u ZDZS 3 m 5 2n 1 3 Howl Wage 19923 Schedule 51992 Figute 4 Change in label foIce paniopatiou women Iquot sk l and group quot1mzv n a MM y Unnamed 1 clnld Unmarue 2 clnldren E MKM igk m a Hm Mamed1cluld Mammy dukhcn s m 1 m2 5 m n 1 Houdv Wag 19925 Schedule 19927 LARGE EHENSIVE MARGIN EFFECTS Figure 5 Change is usual hams if employed 39 310 19 Women by skill an Um a m it v m n Unmmed 1 child Unmaxned children 4 M g a g 2 25 m a Md n Marued clnld Mamed dmmea E g 1 3 n r m s n a a in ii 2n 2 Hourh39 k age 19923 Schedule 5199 Little intensive margin effects Fxgm39e 8 Change in 10g Wages Women by skill and gIonp L mmmed no suds Ummnmd 1 bd Ummned 24 lads 1AM U 1 F W 25 25 Z Named my kids Mnmzd 1 km 399 I ARIpAmv qud U 1 5 m 3 39 1n 3 m1 m 15 m 23 1 aae 19923 Schedule 51992


Buy Material

Are you sure you want to buy this material for

25 Karma

Buy Material

BOOM! Enjoy Your Free Notes!

We've added these Notes to your profile, click here to view them now.


You're already Subscribed!

Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'

Why people love StudySoup

Jim McGreen Ohio University

"Knowing I can count on the Elite Notetaker in my class allows me to focus on what the professor is saying instead of just scribbling notes the whole time and falling behind."

Amaris Trozzo George Washington University

"I made $350 in just two days after posting my first study guide."

Bentley McCaw University of Florida

"I was shooting for a perfect 4.0 GPA this semester. Having StudySoup as a study aid was critical to helping me achieve my goal...and I nailed it!"


"Their 'Elite Notetakers' are making over $1,200/month in sales by creating high quality content that helps their classmates in a time of need."

Become an Elite Notetaker and start selling your notes online!

Refund Policy


All subscriptions to StudySoup are paid in full at the time of subscribing. To change your credit card information or to cancel your subscription, go to "Edit Settings". All credit card information will be available there. If you should decide to cancel your subscription, it will continue to be valid until the next payment period, as all payments for the current period were made in advance. For special circumstances, please email


StudySoup has more than 1 million course-specific study resources to help students study smarter. If you’re having trouble finding what you’re looking for, our customer support team can help you find what you need! Feel free to contact them here:

Recurring Subscriptions: If you have canceled your recurring subscription on the day of renewal and have not downloaded any documents, you may request a refund by submitting an email to

Satisfaction Guarantee: If you’re not satisfied with your subscription, you can contact us for further help. Contact must be made within 3 business days of your subscription purchase and your refund request will be subject for review.

Please Note: Refunds can never be provided more than 30 days after the initial purchase date regardless of your activity on the site.