《財(cái)政學(xué)》（第8版）課件：lecture7 Income redistribution and expenditure programs for the poor

1、Lecture7Income redistribution and expenditure programs for the poorPublic FinanceCurriculum website: http:/jrxynew/kczl/kjzl.jsp?js=all&cou=all2022/7/91Introduction經(jīng)濟(jì)學(xué)家是否應(yīng)該關(guān)注收入分配問(wèn)題？有關(guān)“恰當(dāng)?shù)摹笔杖敕峙溆^念都屬于價(jià)值判斷,沒(méi)有“科學(xué)的”方法來(lái)解決人們?cè)趥惱矸矫娴牟町愐虼擞懻摲峙鋯?wèn)題會(huì)破壞經(jīng)濟(jì)學(xué)的客觀性,經(jīng)濟(jì)學(xué)應(yīng)僅限于關(guān)注效率2022/7/92Introduction但僅僅關(guān)注效率也存在一些問(wèn)題.僅僅關(guān)注效率本身就是

2、一個(gè)價(jià)值判斷問(wèn)題.效率之外還有許多標(biāo)準(zhǔn).決策者總是關(guān)心收入分配的; 如果經(jīng)濟(jì)分析忽略了決策者的關(guān)注，則經(jīng)濟(jì)分析就會(huì)缺乏效果.2022/7/93Distribution of Income首先分析一下家庭之間的收入分配, 我們看看“餡餅”的分配是否公平.表7.1給出了美國(guó)人口普查局公布的自20世紀(jì)60年代以來(lái)部分年度的收入分配數(shù)據(jù).2022/7/94Table 7.1五等分收入法Percentage ShareYearLowest FifthSecond FifthMiddle FifthFourth FifthHighest FifthTop 5 Percent19674.010.817.32

3、4.243.617.219724.744.016.819824.010.016.524.545.017.019873.89.616.124.346.218.219923.89.415.824.246.918.619973.68.915.023.249.421.720023.58.814.823.349.721.720043.48.714.7Source: U.S. Census Bureau, Current Population Survey, Annual Social and Economic SupplementsURL: /hhes/w

4、ww/income/histinc/h02ar.htmlThese figures do not include the value of in-kind transfers.2022/7/95Distribution of Income最富裕20%的人口的收入，占總收入的 50%.而最窮的 20%人口只占總收入的 3.4%.隨著時(shí)間的推移,收入分配的不公平程度提高了.收入差距為什么這么大？家庭收入不公平的最重要原因是戶主的工資差異，而教育的經(jīng)濟(jì)收益提高是工資差異的主要因素2022/7/96Distribution of Income: Poverty貧困線(poverty line): 是足

5、以維持最低生活水平所需要的一個(gè)固定的實(shí)際收入水平.貧困線的確定存在明顯的人為特征，但仍不失為一個(gè)有用的基準(zhǔn)概念.貧困線隨時(shí)間差異而不同不同人群之間也存在差異2022/7/97Poverty Rate (1960-2004)Source: US Bureau of the Census, “Historical Poverty Tables.” WWW Document URL: /hhes/www/poverty/histpov/hstpov2.html2022/7/98Distribution of Income: Poverty2022/7/99Interpretation Problem

6、s對(duì)貧困線 (或貧困率)指標(biāo)的質(zhì)疑有很多.因此，在解讀這一指標(biāo)的時(shí)候，了解形成這一數(shù)據(jù)的傳統(tǒng)方法及其局限性是很重要的.2022/7/910Interpretation Problem #1“收入” 僅僅包括了現(xiàn)金收入.沒(méi)有包括諸如有補(bǔ)貼的醫(yī)療保健或食物劵之類(lèi)的實(shí)物轉(zhuǎn)移(以商品或服務(wù)形式而非現(xiàn)金形式支付給個(gè)人).如果把這部分福利包括在內(nèi)，貧困率會(huì)降低約20%.沒(méi)有考慮非市場(chǎng)服務(wù)的收入，如兒童照料與家務(wù)工作.忽略了耐用品的收入流.2022/7/911Interpretation Problem #2收入指的是稅前的.它忽略了勞動(dòng)所得稅抵免(Earned Income Tax Credit EITC

7、), 這種形式的轉(zhuǎn)移支付在過(guò)去十年高速增長(zhǎng)，現(xiàn)在其規(guī)模已達(dá)到每年 $31billion.這種忽略導(dǎo)致高估貧困率，同時(shí)也影響了其變動(dòng)趨勢(shì).2022/7/912Interpretation Problem #3收入是按年度衡量的.衡量收入的正確時(shí)期并不是明確的.收入在各年之間是有波動(dòng)的.終生收入也許較為理想.考慮一個(gè)正在 “忍饑挨餓”的大學(xué)生 ，從生命周期的角度看，也許他并不是真正的 “窮人”.2022/7/913Interpretation Problem #4觀察單位的界定個(gè)人還是家庭?人們通常作為一個(gè)經(jīng)濟(jì)單位共同作出他們的決策, 家庭比個(gè)人通常更具有規(guī)模經(jīng)濟(jì).Bauman (1997)計(jì)算了

8、一個(gè)結(jié)果：把非家庭成員(如與祖父母同住)的影響考慮進(jìn)來(lái)以后，官方統(tǒng)計(jì)窮困人口的 55%可以剔除.2022/7/914Rationales for Income Redistribution不同種類(lèi)的社會(huì)福利函數(shù)：功利主義極大極小準(zhǔn)則(羅爾斯主義)帕累托效率非個(gè)人主義觀點(diǎn)2022/7/915Simple Utilitarianism功利主義的社會(huì)福利函數(shù)(utilitarian social welfare function)是：它取決于所有n個(gè)社會(huì)成員,一種簡(jiǎn)單的具體函數(shù)形式為：這種具體的函數(shù)形式稱(chēng)為可加的社會(huì)福利函數(shù)(additive social welfare function).202

9、2/7/916Simple Utilitarianism給定可加的SWF, 并假設(shè)：人們具有相同的效用函數(shù)，且只取決于各自的收入.收入的邊際效用是遞減的.社會(huì)的總收入是固定的.據(jù)此可得：政府應(yīng)該再分配收入，以實(shí)現(xiàn)完全公平.2022/7/917Simple Utilitarianism我們用一個(gè)“兩人社會(huì)”來(lái)簡(jiǎn)化說(shuō)明.考察圖 7.1.除I*外的任何收入水平都不可能最大化 SWF.I* 實(shí)現(xiàn)了完全相等的收入.2022/7/918Implications for Income InequalityPauls marginal utility0Peters marginal utilityPauls

10、income Peters income 00MUPaulMUPeteraecdfI*bPauls incomePeters incomeTake ab from Peter and give to PaulPaul gains this much utilityPeter loses this much utilityThis is the net gain to societySocial welfare maximized2022/7/919Simple Utilitarianism一個(gè)激動(dòng)人心的政策含義是：我們應(yīng)當(dāng)追求每個(gè)人完全相同的收入分配政策，然而在認(rèn)同這種政策之前，我們需要謹(jǐn)慎從

11、事.人們是否具有相同的效用函數(shù).收入的邊際效用是否一定遞減.社會(huì)的總收入水平是否一定固定e.g., 旨在進(jìn)行收入再分配的政策往往分減少實(shí)際收入總額.2022/7/920The Maximin Criterion羅爾斯社會(huì)福利函數(shù)(Rawlsian social welfare function)：在這種情況下，社會(huì)福利僅取決于效用最低的那個(gè)人的效用.以其“初始狀況(original position)”觀念為基礎(chǔ)，Rawls (1971)認(rèn)為這一準(zhǔn)則在倫理上具有特殊的正當(dāng)理由.人們對(duì)于他們?cè)谏鐣?huì)中處于收入分配的什么位置并不清楚.2022/7/921The Maximin Criterion對(duì)于

12、這種倫理方面的主張是存在爭(zhēng)議的:對(duì)初始狀態(tài)的自私賦于了特殊的意義人們是極端風(fēng)險(xiǎn)厭惡的一個(gè)能使境況最差人的福利稍有增加，而幾乎所有其他人的福利都變差的機(jī)會(huì)，社會(huì)也應(yīng)該追求嗎？2022/7/922Pareto Efficient Income Redistribution假定富人的效用水平不僅依賴(lài)于其收入，還依賴(lài)于窮人的效用. 也即：這種情況下政府進(jìn)行收入分配可以提高效率，而私人市場(chǎng)卻可能做不到這一點(diǎn)，比如，富人缺乏信息了解誰(shuí)是真正的窮人.可以簡(jiǎn)化為一個(gè)外部性問(wèn)題.2022/7/923Pareto Efficient Income:Redistribution利他主義可以解釋部分政府的收入再分配方

13、案，但并不完全是這樣，比如社會(huì)上的大量慈善捐贈(zèng).此外. 個(gè)人的自利行為扮演重要角色. 假定存在一種不可控因素會(huì)使你變窮.當(dāng)你富有時(shí), 支付 “保險(xiǎn)費(fèi)”. 當(dāng)“不幸”來(lái)臨時(shí)，你就可以收獲 “支付”這也許是某些社會(huì)保險(xiǎn)計(jì)劃的最初動(dòng)機(jī).2022/7/924Nonindividualistic views前面的收入分配觀點(diǎn)認(rèn)為, 社會(huì)福利是個(gè)人效用的某種函數(shù).一些學(xué)者指出：收入分配應(yīng)該獨(dú)立于個(gè)人偏好.比如商品平均主義(commodity egalitarianism). 選舉權(quán), 食物, 住房, 教育, 或者還有保健.2022/7/925Processes versus Outcomes有人認(rèn)為收入的

14、公平分配是由分配的過(guò)程決定的.比如在美國(guó)盛行的 “機(jī)會(huì)公平”.如果機(jī)會(huì)是公平的，那么，不管收入分配的具體情況如何，其結(jié)果都被認(rèn)為是公平的.社會(huì)存在足夠流動(dòng)性的情況下，收入分配不再具有特殊倫理意義 (Gottschalk, 1997).但產(chǎn)生的問(wèn)題是怎么來(lái)評(píng)價(jià)社會(huì)的過(guò)程呢.2022/7/926Expenditure Incidence支出歸宿：支出政策對(duì)實(shí)際收入分配的影響。由于下列原因，支出歸宿難以確定：相對(duì)價(jià)格效應(yīng)（Relative Price Effects）公共品（Public Goods）實(shí)物轉(zhuǎn)移評(píng)價(jià)（Valuing In-Kind Transfers）2022/7/927Relativ

15、e Price Effects假定政府對(duì)低收入者的住房消費(fèi)進(jìn)行補(bǔ)貼.首先想到的是收入由富人轉(zhuǎn)移給了窮人.我們應(yīng)該考慮的是住房補(bǔ)貼的總體效應(yīng)房主獲得了部分補(bǔ)貼轉(zhuǎn)移.建筑行業(yè)的工資與材料價(jià)格也會(huì)上升.通常, 任何政府計(jì)劃都會(huì)引起連鎖價(jià)格變動(dòng)，都會(huì)影響商品消費(fèi)者和投入品供給者的收入，因此最終的歸宿并不清楚.2022/7/928Public Goods窮人與富人能夠從公共品獲得相同收益嗎?確切回答這些問(wèn)題是不可能的，因?yàn)榛谕瑯铀坪跤欣淼募俣ǖ母鞣N答案，具有非常不同的含義.2022/7/929Valuing In-kind Transfers政府通過(guò)提供商品而不是現(xiàn)金來(lái)使窮人獲得收益.食物劵（Food

16、 stamps）公共醫(yī)療補(bǔ)助（Medicaid）公共住房（Public Housing）估算給收益者帶來(lái)的價(jià)值是困難的. 如果轉(zhuǎn)售是困難的，受益人并不總是以一比一的方式來(lái)評(píng)價(jià)實(shí)物轉(zhuǎn)移.考察實(shí)物轉(zhuǎn)移對(duì)效用水平的影響可通過(guò)考察其對(duì)預(yù)算約束的影響2022/7/930In-kind TransfersPounds of cheese per monthOther goods per month30026020150BAD21060FE1UE3420340H2022/7/931In-kind TransfersPounds of cheese per monthOther goods per month

17、30013682150BAD210FE4E5420H1681262022/7/932Numerical Example: BaselineAssume that Jones has the following utility function:Where C indicates the quantity of cheese consumed, and O indicates the quantity of other goods.Jones faces prices PC=2 and PO=1 for cheese and other goods, respectively.What allo

18、cation of would Jones choose with I=300 ?2022/7/933Numerical Example: BaselineThe I=300 means the utility function subject to: 2Pc+Po =300, then:In this Cobb-Douglas utility function, Jones demand curve for cheese is:where the expression came from?2022/7/934Numerical Example: BaselineIn addition, Jo

19、nes demand curve for other goods is:Jones utility is therefore equal to:2022/7/935Numerical Example: Cash TransferIn addition to Jones initial income, assume the government gives a cash transfer of $120.What consumption bundle does Jones now choose, and what is her utility?2022/7/936Numerical Exampl

20、e: Cash TransferIn addition, Jones demand curve for other goods is:Jones utility is therefore equal to:2022/7/937Numerical Example: Binding In-kind Transfer在Jones初始收入的基礎(chǔ)上, 假定政府給他一個(gè)60單位奶酪的轉(zhuǎn)移支付，這60個(gè)單位是不允許轉(zhuǎn)售的.Jones將會(huì)選擇何種消費(fèi)組合, 其效用水平如何?2022/7/938Numerical Example: Binding In-kind TransferNote that the in

21、-kind transfer costs the government $120 (60 units x $2 per unit).When Jones was unconstrained, she used the extra $120 to arrive at an allocation of C,O=52.5,315.Jones cannot attain this, because the minimum amount of C she can consume is C=60 (the amount of the in-kind transfer).2022/7/939Numerica

22、l Example:Binding In-kind TransferThus, she uses all of her fungible income ($300) to purchase the good O:Jones utility is therefore equal to:2022/7/940Numerical Example:Nonbinding In-kindTransfer在Jones初始收入的基礎(chǔ)上, 假定政府給他一個(gè)30單位奶酪的轉(zhuǎn)移支付, 這30個(gè)單位是不允許轉(zhuǎn)售的. 此外，政府還會(huì)再給他$60的現(xiàn)金轉(zhuǎn)移.此時(shí)，Jones的消費(fèi)組合是什么？其效用水平如何?2022/7/9

23、41Numerical Example:Binding In-kind TransferNote that the total transfer costs the government $120. The in-kind transfer costs the government $60 (30 units x $2 per unit), and the cash transfer costs another $60.When Jones was unconstrained, she used the extra $120 to arrive at an allocation of C,O=

24、52.5,315.Jones can attain this, because the minimum amount of C she can consume is C=30, which is less than C=52.5 (the amount of the in-kind transfer).2022/7/942Numerical Example:Binding In-kind TransferThus, she uses part of her fungible income ($360) to purchase the good C and good O.Ultimately,

25、she wants C=52.5, so she purchases 22.5 units of C with her fungible income (with the rest coming from the in-kind benefit).She purchases 315 units of O with the remainder of her fungible income.Utility is the same as the unconstrained case.2022/7/943Valuing In-kind Transfers如果說(shuō)實(shí)物轉(zhuǎn)移會(huì)導(dǎo)致效率低下，它存在的理由是什么

26、呢?商品平均主義（Commodity egalitarianism）減少福利欺詐 (特別是實(shí)物轉(zhuǎn)移中的實(shí)物是低檔品時(shí))因?yàn)閷?shí)物轉(zhuǎn)移也幫助了實(shí)物的生產(chǎn)者，因此在政治上會(huì)更加受到歡迎.2022/7/944Quick Look at Welfare Spending美國(guó)的“福利”是許多不同計(jì)劃的拼湊物.所有的福利計(jì)劃都存在審核問(wèn)題 只有那些低收入者才有資格獲得一些福利計(jì)劃還涉及到家庭結(jié)構(gòu)與財(cái)產(chǎn)等相關(guān)條件福利支出占GDP的比重,過(guò)去30年翻了一倍直接的現(xiàn)金補(bǔ)助減少了，然而健康補(bǔ)貼卻有巨大增長(zhǎng).2022/7/945ProgramFederalState and LocalMedical care$163

27、.8 $118.7 Cash aid82.419.7Food benefits36.82.5Housing benefits34.80.7Education28.81.7Services17.54.7Jobs/training6.90.9Energy aid2.00.1Source: Burke 2003, p. 3. Figures are for 2002.2022/7/946TANF1935-1996: AFDCAid to Families with Dependent Children1996-present: TANFTemporary Assistance for Needy F

28、amilies2022/7/947AFDC/TANF differencesAFDC任何收入低于某一水平且滿足其他條件的人，都可以獲得數(shù)額不等的現(xiàn)金補(bǔ)助沒(méi)有時(shí)間限制沒(méi)有工作要求成本由州與聯(lián)邦政府共擔(dān) 但聯(lián)邦政府沒(méi)有固定限額福利水平的決定受制于聯(lián)邦政府福利受益人掙取收入時(shí)，補(bǔ)助大致一比一地減少TANF這種補(bǔ)助只是臨時(shí)性補(bǔ)助，最多可享受5年有工作要求聯(lián)邦政府給地方政府一筆固定限額的轉(zhuǎn)移支付地方政府對(duì)計(jì)劃的設(shè)計(jì)具有更高的控制力地方政府對(duì)福利受益人的補(bǔ)助與收入掛鉤的政策，沒(méi)有一定之規(guī)2022/7/948Income Maintenance andWork IncentivesFor example,

29、假定某地的福利給付為$300,個(gè)人獲得收入時(shí)的邊際隱含稅率為0.25.G=300 and t=0.25如果個(gè)人掙得 $500, 則其福利給付變化為？收入變化到某一水平時(shí)，個(gè)人將不能再獲得福利給付.$1752022/7/949Income Maintenance andWork Incentives實(shí)際上, 當(dāng)前福利收益B(actual benefit received) 是與隱含稅率t (related to the tax rate), 福利給付 G(welfare grant)和當(dāng)前實(shí)際掙得收入E(actual earnings )的函數(shù).令B=0,有:這里的E稱(chēng)為臨界收入水平：福利收益為

30、零時(shí)的收入水平2022/7/950Income Maintenance andWork Incentives臨界收入水平要求我們?cè)诟＠?jì)劃設(shè)計(jì)時(shí)：較低的稅率t，給福利接受者提供了良好的工作激勵(lì)，但使更多的人有資格獲得福利.例如，如果G=300 and t=0.25, 則能夠獲得福利收益的收入水平為$1,200.而當(dāng)稅率增加到100%時(shí)，能夠獲得福利收益的收入水平為$300.顯然：面對(duì)較高的稅率時(shí)，能夠獲得福利收益的人變少了，但這樣一個(gè)高的稅率會(huì)抑制福利接受者的工作積級(jí)性2022/7/951Analysis of Work Incentives時(shí)間稟賦(time endowment)是指?jìng)€(gè)人可用

31、于勞動(dòng)與休閑的總和.效用函數(shù)由兩種商品組成：leisure and “all other consumption goods” (這里簡(jiǎn)化為收入).U=u(L,C) or 或者等于U=u(L,I) whereL = LeisureC = Consumption goods I = Income2022/7/952w*Analysis of Work IncentivesHours of leisure per monthIncome per month0TaTime EndowmentD|Slope| = wbc2w2022/7/953Analysis of Work IncentivesHo

32、urs of leisure per monthIncome per month0TD|Slope| = wiiiiiiE1FGLeisureWorkIncome2022/7/954Analysis of Work IncentivesHours of leisure per monthIncome per month (= earnings + transfers)0TD|Slope| = wQFGS$100V|Slope| = 3/4wKHours before TANFHours after TANF2022/7/955Analysis of Work IncentivesHours o

33、f leisure per monthIncome per month (= earnings + transfers)0TDPFG$338RP1Budget constraint with t = 100%S0 hours of work selected2022/7/956Analysis of Work IncentivesHours of leisure per monthIncome per month (= earnings + transfers)0TDPMGRE2Hours worked2022/7/957Numerical ExampleAssume that Smith h

34、as the following preferences over leisure and consumption goods:Further, assume that the price of leisure is $5, price of consumption goods is $2, and the time endowment is 100 hours.How much is the leisure demand?2022/7/958Numerical ExampleThe “full” budget constraint is therefore:In general, the d

35、emand curve for good X in a Cobb-Douglas utility function of the form:is2022/7/959Numerical ExampleThis translates easily into leisure demand:She therefore consumes 25 hours of leisure (provides 75 hours of work), earns $375 (=75x$5), and purchases 187.5 units of consumption goods (=$375/2).2022/7/960Introducing the Welfare System into the Analysis這樣一種情況下，他何時(shí)會(huì)失去福利領(lǐng)取資格？ Answer: when her benefits fall to zero, which occurs at the “break