澳大利亞高等教育的成本函數(shù)【外文翻譯】

1、<p>　　本科畢業(yè)論文（設計）</p><p>　　外 文 翻 譯</p><p><b>　　原文：</b></p><p>　　A Cost Function for Higher Education in Australia</p><p><b>　　(extract)</b&g

2、t;</p><p>　　1. Introduction</p><p>　　This study estimates a cost function for higher education in Australian universities. The estimated function is used to examine the cost of providing higher

3、education to overseas students. In the higher education sector, production functions are useful for evaluating the structure of the industry and can serve as guides for individual institutions on policies affecting their

4、 size and scope. Moreover, in view of the relaxation of provisions under which universities can provide higher education to</p><p>　　Cost functions also provide an opportunity to evaluate the cost of providi

5、ng higher education to overseas students. Higher education costs for an individual university can be established from a detailed accounting exercise in which the explicit and implicit costs allocated to courses and level

6、s are collected. However, in obtaining the costs of overseas students, there is still be a need for a cost function because it is necessary to examine what the costs would be under alternative student numbe</p>&l

7、t;p>　　The paper is organised as follows. Section 2 reviews some of the previous work done on the Australian higher education sector and discusses some relevant issues. Section 3 presents the basic cost function model,

8、 describes the data and presents empirical results. Section 4 presents results from the application of the cost function, including the costs predictions for the provision of higher education for overseas students. Secti

9、on 5 draws policy implications and conclusions.</p><p>　　2. Previous Work and Some Issues</p><p>　　Throsby (1986)estimated the cost of providing higher education using data from 18 Australian un

10、iversities for the period 1978 to 1982.This was followed by Lloyd et al.(1993) and Lloyd(1994),which defined the conditions under which the functional form of the cost function allows economies of scale and scope in the

11、production of multiple outputs, that is research and teaching in various disciplines and at different levels. Estimates from these studies were used to analyse the impact of the amalgam</p><p>　　A number of

12、issues arise in deriving and using such a model. First, previous studies failed to account for unobserved differences between universities. Each institution has a particular structure that may make the provision of parti

13、cular courses more cost efficient in one university than in others. In addition, the explanatory variables chosen may not adequately represent the characteristics of particular institutions; for example, there may be var

14、iation of quality and the proportion of resource</p><p>　　Second, there are aggregation problems associated with the accurate specification of the variables. These occur because the choice of discipline or c

15、ourse level categories masks large differences between components of the aggregated discipline. For example, health includes the training of both high-cost doctors and relatively low-cost nurses. At the aggregate level t

16、hey may be combined into a single composite group, but individual universities may specialise in either the high-cost or the low-co</p><p>　　Third, the conventional cost minimisation assumption does not appl

17、y naturally to the higher education sector when government (to a large extent)determines both the funding and the output. In fact, many universities may use the same model as used by the government to determine the fundi

18、ng of domestic students (the Relative Funding Model)for setting fees for overseas students.</p><p>　　In general, the first and second problems are features of the availability of appropriately disaggregated

19、data and model specification. Preliminary investigation has revealed that while finding an answer to the second problem is difficult, a more rigorous study can tackle the first. This third problem is intrinsic to the cos

20、t function methodology. </p><p>　　The cost function approach is nonetheless a useful tool for higher education policy development. Universities can compare their own performance with the average performance

21、estimated by the cost function. Universities can vary size and structure of the student body and consider the likely effect on their budget. The cost function may also be fed into the mechanism for providing funding for

22、domestic students through the Relative Funding Model. The cost function may also be used to evaluate new wa</p><p>　　3. (ellipsis)</p><p>　　4. Application of the Cost Function</p><p&g

23、t;　　4.1. Total Costs</p><p>　　In this subsection, the estimated coefficients of the chosen cost function above are used to derive estimates of total cost as defined in equations(2).The predicted values for 1

24、997,using data for all students, are presented in Tables 3 and 4 below. In Table 3,the universities are arranged from smallest to largest(by EFTSU load),and are presented with information on student load. Total cost figu

25、res–both actual and predicted–are also shown. The model yields predicted values of total costs that are</p><p>　　The cubic form of the cost function indicates rising costs of teaching at low student load lev

26、els, while the rate of growth gradually diminishes as student numbers rise because of gains from economies of scale. At very high student numbers, the cost function implies a renewed rapid rise in the costs incurred by A

27、ustralian universities. A typical cubic cost function is presented in Figure 1,which represents the cost of providing higher education to a Postgraduate by Research student, using UNSW f</p><p>　　The total c

28、ost curves for Postgraduate by Coursework Students, using fixed values for all the other variables, yield a flatter cubic curve than for postgraduate by research work. Whereas most students undertaking postgraduate by re

29、search degrees are PhD students, the coursework postgraduate courses are predominantly at the Masters level. Masters level courses would not require as expensive a staff profile as PhD students.</p><p>　　A c

30、ontrast is shown by the corresponding curve for Undergraduate students, shown in Figure 2, where costs monotonically increase by virtue of the positive coefficients obtained for all three undergraduate student load regre

31、ssors,L3,L3 andL3 ,and over the range shown the cubic term appears to have little effect. In this case, economies of scale gained from increasing postgraduate students numbers are not realised in the same way with the in

32、crease in undergraduate student numbers. It is also possi</p><p>　　4.2. Average and Marginal Costs</p><p>　　The average and marginal cost estimates for 1997 are presented in Table 4.The average

33、cost of higher education across the universities ranged between between$7900 to just under $14000 in 1997.Columns (3)to(5)present the marginal costs by level of study. The marginal costs for undergraduate students are ar

34、ound$2000 for smaller universities and these costs tend to increase with the EFTSU load of the universities. Monash University, the largest university with 15 times the student load of the NTU(t</p><p>　　Mar

35、ginal costs of coursework graduate students lie in between those of the undergraduate and research postgraduate students, but are also positively correlated with the size of the university. In contrast, the marginal cost

36、s for Research Postgraduate students are substantially higher than both Coursework Graduate and Undergraduate students, and university size does not explain the observed variations. Of the ten largest universities on the

37、 list, three have marginal costs over$75,000 for their Re</p><p>　　In general, the marginal cost of higher education is U-shaped, as in Figure 3 below, reflecting increasing returns with economies of scale,

38、then diminishing returns derived from fixed inputs to teaching. At low student numbers, marginal costs are relatively high but this gradually declines as student numbers rise to about 1200 EFTSUs. Thereafter, the margina

39、l costs increase.A typical marginal cost curve for Postgraduate Research students is shown in Figure 3, derived using fixed values for postg</p><p>　　The cost of providing postgraduate courses is higher than

40、 undergraduate courses because of the high level of specialised resources required. The distribution of students across the various disciplines is also important in explaining differences between Universities, with many

41、universities specialising in the provision of certain courses.</p><p>　　4.3. The Cost of Overseas Students</p><p>　　The cost function can be used to estimate the cost of providing higher educati

42、on to overseas students. However, information on the costs of higher education in Australia is available only for the total number of students, that is, overseas plus domestic students, at a particular date. No data are

43、available relating to the cost that would be incurred if only domestic students were being educated. This means that estimates of the extra cost involved in educating overseas students can only be obtai</p><p&

44、gt;　　The cost function is applied to domestic student data to predict the average and marginal costs for domestic students. Corresponding values for overseas students are obtained as the residual values of the cost estim

45、ates for all students and for domestic students only. The results of this exercise for 1997 are shown in Table 5.</p><p>　　Universities across Australia vary widely in the distribution of student load betwee

46、n domestic and fee-paying overseas students. In general, the overseas student population comprise less than five per cent of the student population in most regional universities, but are much more important in universiti

47、es located major city centres .In 1997,the top five universities, in terms of numbers of full-fee paying foreign student loads are Monash University, RMIT, the University of Melbourne, UNSW and Cu</p><p>　　T

48、here is a compositional issue in generating costs for overseas students. If the cost structure of overseas students is projected from information about all students and domestic students, the implicit assumption is made

49、that the discipline composition of overseas students is the same as that of domestic students. However, this is not the case. Data from DETYA,1998 show the discipline composition of overseas students. In 1997 about half

50、of all overseas students were undertaking courses in the Bu</p><p>　　With this caveat in mind, the average costs are shown in the final column of the table. In general, the apparent average cost of higher ed

51、ucation provision for overseas students is greater than for domestic students. In many institutions, these apparent cost differences are greater by a factor of two or three. The table also shows that while the estimated

52、costs of higher education for domestic students are comparable across the universities nationwide, the corresponding costs for overseas studen</p><p>　　Across universities, the per capita cost of teaching fo

53、r a domestic students range only between$7000 and$12000, reflecting the discipline and level of course structure of the university. In contrast, the apparent per capita cost of teaching overseas students varies widely,fr

54、om$9300 to as much as$48,500.The two smallest universities, the Northern Territory University and the University of Ballarat, incur the highest apparent per capita cost in mounting courses for overseas students of at lea

55、st$40</p><p>　　A number of universities incur apparent per capita cost for overseas students between $20,000 and$29,000.These include Macquarie University, Sydney University of Technology, UNSW, University o

56、f Queensland and the University of Sydney. The relatively low per capita costs incurred by RMIT and Curtin for its overseas students (with large foreign student shares in their total student population)may be a result of

57、 two effects. First, economies of scales are achieved with relatively high student numbe</p><p>　　5. Conclusion</p><p>　　Information derived from the estimation of a simple cost function can ass

58、ist planners and university managers in addressing problems of resource allocation and pricing. Although methodological difficulties exist, the estimation of a cost function can nevertheless help to view university produ

59、ction and cost relationships in a systematic and theoretically plausible way. </p><p>　　In this study, new cost functions were estimated for the Australian university sector using pooled series of cross-sect

60、ional data. The study enabled the quantification of the cost differences between levels of studies and subject areas. It was hypothesised that university total costs would exhibit varying cost movements along production

61、stages suggesting a higher order functional form. Accordingly a cubic functional form was used to estimate model. It was found that the cubic and quadratic terms</p><p>　　The new cost function will be a usef

62、ul tool for university managers in exploring the effect of varying size and structure of courses on university budgets. To illustrate the use of the tool the model was used to derive estimates of the overall cost of prov

63、iding higher education to overseas students. Application of the model suggested a large variation in average costs per overseas student between universities reflecting student load and course composition.</p><

64、p>　　文章出處：Discussion Papers in Economic, Finance and International Competitiveness No175 ，P1-P12，P18</p><p><b>　　譯文：</b></p><p>　　澳大利亞高等教育的成本函數(shù)</p><p><b>　　(節(jié)選)

65、</b></p><p><b>　　引言</b></p><p>　　本文主要評價澳大利亞高校教育成本函數(shù)，運用這個函數(shù)來計算海外學生接受高等教育的成本。在高等教育部門，生產(chǎn)函數(shù)有利于評價產(chǎn)業(yè)結構，而且能夠影響私人機構在政策引導上的大小和范圍。此外，鑒于高校為國內學生提供全額支付的教育政策上的放松，成本函數(shù)有助于制定合理的定價政策。</p>

66、<p>　　成本函數(shù)也為評估海外學生接受高等教育的成本提供了機會。私立高校的教育成本是建立在詳細的核算基礎上的，在核算過程中，將分配到課程和各階層的顯性和隱性成本歸集起來。但是，為了準確計算海外學生的教育成本，仍然需要成本函數(shù)。因為有必要評估在有和沒有海外學生情況下的總學生數(shù)的成本。對于不同學生成本變動的情況可以通過高校間同一時期不同學生的比較來取得。因此成本函數(shù)的使用為計算海外學生接受高等教育的凈收益（或成本）提供了方法。

67、</p><p>　　本文由以下部分構成。第二部分回顧了澳大利亞高校教育部門先前所做的工作，并且研究了一些相關的重要問題。第三部分列出了主要的成本函數(shù)模型以及相關數(shù)據(jù)和結果。第四部分列出了運用成本函數(shù)所得到的結論，包括海外學生接受高等教育的成本預測。第五部分得出政策涵義和結論。</p><p><b>　　前期工作和相關問題</b></p><p&

68、gt;　　Throsby(1986)估算的高校教育成本所使用的數(shù)據(jù)來自1978年到1982年間的18個澳大利亞大學。后來，Lloydetal.(1993) 和Lloyd(1994)明確在成本函數(shù)的形式下，能夠節(jié)省產(chǎn)品多樣化輸出的規(guī)模和范圍，這是從多學科和不同階層下研究得出的。這些研究都用來分析1987年改革形成的融合所產(chǎn)生的影響。特別是這個計算方程用來估算兩個或兩個以上機構混合所形成的可能的節(jié)約成本。澳大利亞最新的成本函數(shù)是由Thros

69、by 和 Heaton(1995)得出的。他們運用1991年42家機構的交叉數(shù)據(jù)分析了機構運營成本和10個廣闊的學科領域和3個研究層次的學生數(shù)量之間的關系，同時利用學生數(shù)量的二次方函數(shù)（包括邊際成本）。他們的成本函數(shù)被Baker, Creedy 和 Johnson(1996)用來計算海外學生的教育成本。</p><p>　　同時，在運用此模型時也產(chǎn)生了一系列問題。首先，前面的研究沒有解釋高校間的差異。高校中各機構

70、有特殊的結構，這些機構的一些特別的課程可能會產(chǎn)生不同于其他學校的成本效益。而且，變量的選擇也可能不符合這些機構的特點。例如：會出現(xiàn)質量以及研究資源的比例的變化。因此需要作出一個假設就是教學質量和學生數(shù)量的變化無關。</p><p>　　其次，是關于對變化準確描述的相關問題。這些問題是因為對學科和課程水平分類的選擇而隱含在各學科組成部分中的差異所引起的。就像醫(yī)療水平也包括了對高成本醫(yī)生和相對低成本護士的培訓。在總體

71、水平上，他們可能由單個的團隊組成，但是私立大學可能要么研究高成本要么研究低成本的產(chǎn)品。因此，他們的總成本會因為將大學作為一個整體而使用一個平均參數(shù)被高估或低估。</p><p>　　第三，傳統(tǒng)成本最小化的假設不能自然運用到各個高校，即使政府能夠保證資金的輸出。實際上，很多高?？赡芨鶕?jù)政府保障的國內學生的資金來運用相同的模型制定海外學生的學費。</p><p>　　一般來說，第一和第二個問題

72、的特點是集中在對分散數(shù)據(jù)的合理使用和模型敘述上。初步調查發(fā)現(xiàn)找到第二個問題的解決方法有些困難，而這項更嚴密的調查能夠解決第一個問題。第三個問題涉及內在的成本函數(shù)方法論。</p><p>　　成本函數(shù)方法仍然是高校教育政策發(fā)展的有效工具。高校能夠通過成本函數(shù)來比較自身績效與平均水平間的差異，從而調整規(guī)模大小和學生構成以及考慮預算的影響。成本函數(shù)通過相關輔助模型也有助于完善對國內學生提供資金的機制。成本函數(shù)也用于評估

73、補償機制的新方法中以及對國內全額交納學費學生的定價政策。它也可以估算為海外學生提供高等教育的成本和收益，從而指導相應的定價政策。</p><p><b>　　3、（略）</b></p><p><b>　　4、成本函數(shù)的應用</b></p><p><b>　　4.1、總成本</b></p>

74、;<p>　　這一部分中，上述選擇的成本函數(shù)的估算系數(shù)是用來估算等式（2）中確定的總成本。1997年的預測值列示在下列表格3和4中。表格3中高校從小到大排列列示了學生貸款數(shù)，也列示了實際和預測的總成本。模型得出了總成本的預測值，這個值合理接近實際值，70%的預測值中有10%的錯誤限度。這也表明了在學生貸款和教學總成本之間相對固定的關系，也暗示了平均成本所在的一個不太寬松的邊緣。這個問題會在下面進一步研究。</p>

75、;<p>　　成本函數(shù)的立方式反映了較低的學生貸款水平下增長的教育成本，同時，增長率隨著學生人數(shù)的上升逐漸下降，這是由于經(jīng)濟規(guī)模所獲得的利潤引起的。在學生人數(shù)較多的高校，成本函數(shù)顯示成本正不斷增加。一個典型的三次成本函數(shù)在表格1中列示，它代表了研究生的教育成本，同時使用了各學科和課程水平的固定價值。這個曲線是澳大利亞進行研究生教育的大學的標準。對于小型的高校，總成本曲線也是相同形狀，只是在相對較低的成本水平。研究生貸款的增

76、加可能是由于高校不斷增加的課程數(shù)目帶來的越來越昂貴的員工測評而導致的。隨著高校博士項目的增加，就有可能增加更加昂貴的研究項目數(shù)量，同時這也會引起教職工更高的報酬要求。</p><p>　　通過使用其他變量的固定價值，研究生的總成本曲線更加平緩。然而大多從事研究領域的研究生都會成為博士生，而他們在碩士生階段的課程作業(yè)表現(xiàn)更優(yōu)。碩士階段所需的員工報酬卻沒有博士階段高。</p><p>　　通過

77、與研究生相對應的曲線的對比，表格2中顯示，成本隨著研究生貸款額增加而單調遞增，所有形式中，三次方形式只有很小的影響。因此，經(jīng)濟效益并不與因為研究生數(shù)量的增加而得以實現(xiàn)。高校的研究層次中可能會產(chǎn)生教育成本的交叉補助，這會讓一些具有研究生的特色研究課程更便宜。</p><p>　　4.2、平均成本和邊際成本</p><p>　　表格4顯示的是1997年的平均和邊際成本。1997年高校教育的平均

78、成本是從7900美元到14000美元。柱狀圖3和5顯示了個各學科層次的邊際成本。小型高校研究生的邊際成本大約在2000美元，而且隨著EFTSU貸款而增加。Monash大學，它擁有的學生是NTU大學（最小大學）的15倍，它的研究生的邊際成本是16300美元。一般來說，研究生的邊際成本比本科生要小。</p><p>　　本科在讀生的邊際成本存在于本科畢業(yè)生和研究生兩者中，但是也的確與學校規(guī)模有關。相反，研究生的邊際成

79、本實際上要高于本科畢業(yè)生和在讀生，而且學校規(guī)模也不能解釋其中的差異。表中10個規(guī)模最大的高校中，有三所的研究生邊際成本超過75000美元，但同時，剩余學校的邊際成本在35000美元以下。在另一個級別的大學中，一些小規(guī)模學校的邊際成本超過60000美元。一些高校在讀本科生的邊際成本明顯高于研究生的邊際成本：這些學校包括Macquarie 大學、La Trobe大學 、RMIT、 QUT大學和Monash大學。這些差異是由一系列的原因所致。

80、首先，大型高校沒有必要一定需要大量的研究生數(shù)量。第二，各高校間特色的高成本課程的構成存在很大不同。第三，一些小規(guī)模大學的加入可能產(chǎn)生復雜的結果。</p><p>　　一般而言，教育邊際成本呈U字形，正如圖3所示，邊際成本隨著資金規(guī)模的增加而減小，到最低值后又隨著資金規(guī)模的增加而增加。學生數(shù)量較少時，邊際成本相對較高，但會隨著學生數(shù)量的增加到大約1200 EFTSU而逐漸下降。此后邊際成本逐漸增加。圖3所示的這條重

81、要的邊際成本曲線是根據(jù)本科生和研究生貸款數(shù)量得到的。</p><p>　　研究生的教育成本由于對特別資源的更高要求而高于本科生。由于制度差異而導致的學生分配情況對分析高校間的差異有重要意義，因為一些高校在某些課程的選擇上是由自身特點所決定的。</p><p>　　4.3、海外學生成本</p><p>　　成本函數(shù)能夠運用到估算海外學生高等教育成本中，但是在澳大利亞，

82、有關教育成本的信息只能針對總體學生數(shù)，也就是某一時點國內學生和國外學生之和，沒有相關數(shù)據(jù)時是只針對國內學生教育成本的。這意味著其他包括外國學生教育成本的估算只能通過成本函數(shù)和估算其他關于學生數(shù)量下的成本來獲得。因此這種估算的質量只能通過經(jīng)驗得來的成本函數(shù)來決定。</p><p>　　成本函數(shù)利用國內學生的數(shù)據(jù)來預測國內學生的平均成本和邊際成本。相應的海外學生的價值通過全體學生和國內學生的成本的剩余價值來獲得。由此

83、得出的結果在表格5中顯示。</p><p>　　澳大利亞高校對于國內學生和海外學生間貸款的分配有很大差異。一般來說，很多高校的海外學生人數(shù)比總人數(shù)少5%，但在位于城市中心的大學里他們是很重要的一部分。1997年，根據(jù)大學全額支付學費的海外學生貸款數(shù)量，排名前五名的大學是Monash大學、RMIT、墨爾本大學、UNSW和Curtin技術學校。但是就相對的學生份額，RMIT海外學生貸款占了29%，是最高的，其次是Cu

84、rtin技術學校，占了27%。</p><p>　　海外學生產(chǎn)生的成本有它的構成條件。如果海外學生的成本構成來自全體學生和國內學生的信息，就要做出一個隱含著的假設，就是海外學生的構成原則與國內學生相同。但是，這不是原因。DETYA的數(shù)據(jù)顯示了海外學生的構成原則。1997年，有一半的海外學生主修商業(yè)、管理、經(jīng)濟學等課程，而國內學生主修這些課程的人數(shù)只有20%。因為商業(yè)、管理、經(jīng)濟學是相對低成本的課程，所以上述假設可

85、能會使海外學生的平均成本膨脹。</p><p>　　由此而產(chǎn)生的平均成本可以在柱狀圖中表示。一般來說，海外學生高等教育的顯性平均成本高于國內學生。這些顯性成本差異是有兩種或三種原因造成的。圖標顯示，高校間國內學生的高等教育成本是可以比較的，而海外學生相應的成本變化很大，不易比較。</p><p>　　各個高校中國內學生的每單位教育資本成本從7000美元到12000美元，它反映了高校各學科和

86、課程的結構。相反，海外學生的教育成本差異很大，從9300美元到48500美元。其中兩所規(guī)模最小的大學，Territory大學和Ballarat大學，他們承擔的最高的資本成本至少有40000美元。</p><p>　　一些高校承擔的海外學生每單位顯性資本成本在20000美元和29000美元之間。他們包括Macquarie大學、悉尼技術學院、UNSW、昆士蘭大學和悉尼大學。而RMIT和Curtin大學所承擔的相對低的

87、資本成本是有兩點引起的。首先，經(jīng)濟規(guī)模由相對較多的學生形成。其次是在總的貸款中海外學生所占的比例較大，上述提到的構成問題也不是很重要，因為海外學生的制度結構和全體學生的制度結構很接近。</p><p><b>　　5、結論</b></p><p>　　從一個簡單的成本函數(shù)得到的信息有助于計劃制定者和高校管理者解決資源分配和定價的問題。雖然還有技術層面的困難存在，但成本

88、函數(shù)的計算在合理的系統(tǒng)和理論化方式下對了解高校的產(chǎn)出和成本間的關系有很大幫助。</p><p>　　在研究中，新的成本函數(shù)也適用于使用合并交叉數(shù)據(jù)的高校部門。研究層面和學科領域之間的成本定量也有差別。同時假設了高校的總成本會反映出隨著更高效的產(chǎn)出而帶來的成本變動。因此估算模型中使用了三次方的形式。研究發(fā)現(xiàn)，大多數(shù)的實例表明二次方和三次方的形式都有助于計算總成本。</p><p>　　新的成