【正文】 eral channels through which inequality might affect growth, the potentially most important of which are briefly surveyed hereInequality, saving, and consumptionThe traditional theory about the link between ine distribution and growth is that greater inequality leads to greater savings, and thus greater capital accumulation and growth. Kaldor (1957) assumes that all wages are consumed and only profits provide resources for additional capital accumulation. In Luigi Pasinetti (1962), this is interpreted as a consumption function in which the rich have a lower marginal propensity to consume than the poor, so that increasing in equality will result in higher savings and thus more rapid capital accumulation and growth. Time has not served this hypothesis well. Since the pioneering contributions of Solow (1957), we no longer think of capital accumulation as being the most important determinant of growth. Still, in the Solow model, capital accumulation is the single most important factor that countries have wellestablished means of influencing—although these means do not include distributional policyThe KaldorPasinetti channel is grounded in applying a Keynesian style aggregate consumption function across households. But, at virtually the same time that these ideas were being developed, Milton Friedman (1957) showed that this consumption function was the result of confusing total ine with permanent ine. If the transitory ponent of ine were removed from consumption, Friedman found that savings was proportional to (permanent) ine。n 1997).Imperfect information and agency costsThe second mechanism by which ine distribution can affect growth es from the economics of information. In the presence of inplete information, the Second Fundamental Theorem of Welfare Economics will not hold, and there will be some government interventions that represent Pareto improvements over the petitive equilibrium (Greenwald and Stiglitz 1986). This general theorem has specific applications to inequality. Inequalities in wealth will generally require the owner of assets, the “principal,” to delegate the use of assets to another, the “agent.” In the presence of inplete information, it will be impossible to write down a plete contract to specify what actions the agent should undertake in each contingency or at least impossible to monitor (and thus enforce) that contract. As a result, a principalagent problem arises, creating what are sometimes called “agency costs.” Agency costs and the steps taken to mitigate them—like engaging in costly monitoring—affect efficiency in a number of ways. If the extent of inequality affects the extent of agency costs, then it will also affect the overall efficiency of the economy.Perhaps the most obvious example of this es from developing countries,where sharecropping contracts are a widespread way to solve the problem of costly monitoring by ensuring that the tenant has an incentive to by imposing,in effect,a marginal(and average)tax rate of between 33 percent and 50 percent,these contracts lead to undersupply of effort and underinvestment in the land. If a more equitable distribution of land allowed farmers to buy their land,the result could be an improvement in efficiency.(Jason Furman,Joseph E. Stiglitz. Ine Distribution and Growth[J].The America Economic Review,1999,56:225230.)收入分配和增長賈森斯蒂格利茨 著 鄧娜 譯傳統(tǒng)上,總體經(jīng)濟活動和收入分配之間關(guān)系的研究強調(diào)從輸出到收入分配的鏈接。但是，更全面的研究發(fā)現(xiàn)，庫茨涅茨曲線在數(shù)據(jù)上幾乎沒有根據(jù)。戴尼爾和林在大多數(shù)的案例中，不平等的變化相對較小，且與初始收入無關(guān)。另外還存在一個追隨劉易斯（1954年）和尼古拉研究確定了不平等可能影響收入增長的幾個渠道，在此對潛在的最重要的渠道進行了簡要的研究?？柖啵?957）假設(shè)所有的工資都消耗了，只有利潤提供額外資本積累的資源。帕西內(nèi)蒂模型中（1962），這被解釋為一個消費函數(shù)，其中富人有比窮人更低的邊際消費傾向，因此，不平等的增加將導(dǎo)致更高的儲蓄，從而導(dǎo)致了更迅速的資本積累和增長。因為索洛的開創(chuàng)性貢獻（1957年），我們不再認為資本積累是經(jīng)濟增長的最重要因素?？柖嗯廖鲀?nèi)蒂渠道，主要建立在凱恩斯主義風(fēng)格應(yīng)用在家庭總消費的功能基礎(chǔ)之上的。他發(fā)現(xiàn)，如果暫時性收入從消費中剔除了，那么儲蓄以固定的邊際消費傾向與（永久性）收入成正比，收入分配對儲蓄沒有影響。勞倫斯薩默斯(1981)認為財富的很大一部分原因是沒有生命周期儲蓄。（這與不平等程度高的調(diào)查結(jié)果一致，大于可占的生命周期儲蓄本身的）由于凸儲蓄功能，不平等程度的增加將增加儲蓄。東亞的經(jīng)驗令人信服地表明，即使貧窮的農(nóng)民也可以有很高的儲蓄率。不完全信息和代理成本收入分配會影響經(jīng)濟增長的第二個機制來自信息經(jīng)濟學(xué)。財富的不平等一般需要資產(chǎn)所有者，即“委托人”，將資產(chǎn)的使用權(quán)交給另一個人，即 “代理人”。因此，一個委托代理問題的提出，產(chǎn)生了“代理成本”。如果不平等程度的大小影響了代理成本的大小，那么它也會影響經(jīng)濟的整體效率。但實際上卻迫使邊際稅率和平均稅率在33%~50%之間，這些合同將導(dǎo)致在土地方面的供應(yīng)和投資不足。(賈森[J]. 美國經(jīng)濟評論，1999，56：2252