1992年--統(tǒng)計學外文翻譯---測度1980-1989間瑞典制藥業(yè)生產率變化基于非參數(shù)的malmquist指數(shù)法

1、<p>　　4360漢字，2700單詞，1.4萬英文字符</p><p>　　畢業(yè)論文(設計)外文翻譯</p><p>　　學 院：統(tǒng)計與數(shù)學學院</p><p>　　專 業(yè)：統(tǒng)計學</p><p><b>　　班 級：統(tǒng)計</b></p><p><b>　

2、　學 　 號：</b></p><p><b>　　學生姓名：</b></p><p><b>　　指導教師： </b></p><p>　　二○一五 年 三 月</p><p><b>　　外文翻譯之一</b></p><p>　　Prod

3、uctivity Changes in Swedish Pharmacies 1980-1989：A Non-Parametric Malmquist Approach</p><p>　　Author(s)：R. FÄRE; S. GROSSKOPF; B. LINDGREN; P.ROOS</p><p>　　Nationality：USA；Swedish</p>

4、;<p>　　Source：The Journal of Productivity Analysis, 3, 85-101 (1992)</p><p><b>　　Abstract </b></p><p>　　In this article we develop a non-parametric (linear programming) approa

5、ch for calculation of a Malmquist (input based) productivity index. The method is applied to the case of Swedish pharmacies.</p><p>　　1. Introduction</p><p>　　The purpose of this article is to d

6、evelop an input based non-parametric methodology for calculating productivity growth and to apply it to a sample of Swedish pharmacies. Our methodology merges ideas from measurement of efficiency by Farrell [1957] and fr

7、om measurement of productivity as expressed by Caves, Christensen, and Dicwert [1982]. In his classic article, “The Measurement of Productive Efficiency,” Farrell introduced a framework for efficiency gauging in which ov

8、erall efficiency can be d</p><p>　　In Sweden, the retail trade of pharmaceutical products has been the responsibility of a public monopoly since 1971. In their agreement with the Swedish Government, Apoteksb

9、olaget (the National Corporation of Swedish Pharmacies) “is responsible for ensuring that an adequate supply of drugs is maintained in the country. For this purpose, the business shall be conducted to foster opportunitie

10、s for taking advantage of pharmaceutical advances, while maintaining drug costs at the lowest possible level</p><p>　　At present, Apoteksbolaget calculates the productivity of a pharmacy as the ratio between

11、 a weighted sum of four outputs and the total number of hours worked for two categories of personnel. The weights assigned to the respective outputs are assumed to reflect differences in resource use, and in the calculat

12、ion of total output aggregate, all pharmacies are assigned the same weights. Total labor input is obtained as the sum of hours worked by the two types of personnel. For each pharmacy, product</p><p>　　Our sa

13、mple consists of 42 group (or regional) pharmacies operating in Sweden from 1980 to 1989. These group pharmacies are a small part of the total number of pharmacies in Sweden (there were 816 pharmacies in 1989). We focus

14、on these 42 groups pharmacies for several reasons. First, we had data on these 42 over the entire time period. Second, the fact that they are all group pharmacies (as opposed to local or hospital pharmacies) means that t

15、heir responsibilities and sizes are fairly similar. Th</p><p>　　Relative to the method presently used by Apoteksbolaget, our method is different with respect to the degree of productivity change, as well as

16、with respect to direction of change in some cases. For the sample as a whole we find productivity increasing in seven periods and productivity declining in two. The method presently used by Apoteksbolaget also yields pro

17、gress in seven periods and regress in two. However, the important point to observe is that periods with progress/regress were not always </p><p>　　The method of calculating productivity and productivity chan

18、ges presently used by Apoteksbolaget has many drawbacks, e.g.: (1) It assumes that the underlying pharmacy technology is of a very special form (which may not be an appropriate assumption for the pharmacy production tech

19、nology); (2) It cannot distinguish between changes in efficiency and change in the frontier technology; (3) It cannot easily include more input variables other than labor and requires outputs to be measured in the same

20、</p><p>　　2. The productivity index</p><p>　　The production technology is defined at each period ,，to be the set of all feasible input and output vectors. If denotes an input vector at period a

21、nd an output vector in the same period, then the technology is the set , where . We also model the technology by the input correspondence or equivalently by the input requirement set</p><p><b>　　(1)&l

22、t;/b></p><p>　　The input requirement set , denotes all input vectors capable of producing outputs during period . Here we assume that is a closed convex set for all , and that there is no free lunch, i

23、.e., . Moreover, we impose disposability of inputs and outputs, i.e., and , respectively.</p><p>　　In this article, we formalize equation (1) as a piecewise linear input requirement set or equivalently as an

24、 activity analysis model. The coefficients in this model consist of observed inputs and outputs. We assume that there are observations of inputs in each period . These inputs are employed to produce of observed out

25、puts, , at period , and we assume that the number of observations are the same for all , i.e., .</p><p>　　The input requirement set (1) is formed from the observations as (see Färe, Grosskopf, and Lovel

26、l [1985])</p><p><b>　　(2)</b></p><p>　　where is an intensity variable familiar from activity analysis. The intensity variables serve to form technology, which here is the convex con

27、e of observed inputs and outputs. Constant returns to scale is imposed on the reference technology, but other forms of returns to scale may be imposed by restricting the sum of the intensity variables (see Grosskopf [198

28、6]). One may also show that satisfies the properties introduced above (see Shephard [1970] or Färe [1988]).</p><p>　　The Malmquist input based productivity index is expressed in terms of four input dis

29、tance functions. The first is defined as</p><p><b>　　(3)</b></p><p><b>　　Clearly,</b></p><p>　　as the following figure 1 illustrates.</p><p>　　I

30、n figure 1, the input vector belongs to the input requirement set . The distance function measures the largest possible contraction of under the condition that is feasible, i.e., . In terms of figure 1, . For observa

31、tion , , the value of the distance function is obtained as the solution to the linear programming problem</p><p><b>　　(4)</b></p><p>　　Note that is an element of the input set whic

32、h implies that the distance function takes values larger than or equal to one. The value one is achieved whenever the input vector belongs to the isoquant of the input set, and hence where it is technically efficient &#

33、224; la Farrell [1957].</p><p>　　Figure 1. The input distance function.</p><p>　　We note the input distance function is the reciprocal of the Farrell technical efficiency measure, a fact which w

34、e have exploited to calculate the distance function.</p><p>　　In order to define the input based Malmquist productivity index by Caves, Christensen, and Diewert [1982], we need to relate the input output vec

35、tors at period to the technology in the succeeding period. Therefore, we evaluate the input distance function for an input output vector at period relative to the input requirement set in the following period.</

36、p><p><b>　　(5)</b></p><p>　　Again, . However, need not be feasible at , thus if equation (5) has a solution (i.e., supremum is a maximum), the value of may be strictly less than one.&

37、lt;/p><p>　　In our data set, the observed input , is positive for each observation and each period. This together with strong disposability of inputs and constant returns to scale ensure that we can calculate t

38、he value of the input distance function (5) for , , as the solution to the linear programming problem</p><p><b>　　(6)</b></p><p>　　We note that since need not be a member of the inp

39、ut requirement set , the value of this distance function may be strictly less than one.</p><p>　　Two additional evaluations of the input distance function are required in order to define the productivity ind

40、ex. We need to evaluate observations at relative to the technologies at and . In particular,</p><p><b>　　(7)</b></p><p><b>　　and</b></p><p><b>　　(8)&

41、lt;/b></p><p>　　The computation of equation (8) is identical to that of equation (3) so that in equation (4) we need only substitute for . The computation of equation (7) is parallel to that of equation

42、(5), and again we need only substitute for and vice versa. We note of course that since need not be feasible under the technology , the input distance function may be strictly less than one.</p><p>　　Fol

43、lowing Caves, Christensen, and Diewert [1982], we define the input based Malmquist productivity index as</p><p><b>　　(9)</b></p><p>　　Actually, our definition is the geometric mean o

44、f two Malmquist indexes as defined by Caves, Christensen, and Diewert [1982].</p><p>　　In their work, Caves, Christensen and Diewert [1982] make two assumptions. First, they assume that and equal unity for

45、each observation and period. In the terminology of Farrell [1957], this means that there is no technical inefficiency. Second, they assume that the distance functions are of translog form with identical second order term

46、s. Here we follow Färe et al. [1989], and model the technology as piecewise linear and allow for inefficiencies. By allowing for inefficiencies, the productivit</p><p><b>　　(10)</b></p>

47、;<p>　　where the quotient outside the bracket measures the change in technical inefficiency and the ratios inside the bracket measure the shift in the frontier between periods and as figure 2 illustrates.</p&

48、gt;<p>　　We denote the technology at by and at by , and note that and that is similarly defined. The two observations and are both feasible in their respective periods. We may express the productivity index i

49、n terms of the above distances along the x-axis as</p><p><b>　　(11)</b></p><p>　　where denotes the ratio of the Farrell measure of technical efficiency and the last part is the geom

50、etric mean of the shifts in technology at and . Note that the shifts in technology are measured locally for the observation at and . This implies that: 1) the whole technology need not behave uniformly, and 2) that tec

51、hnological regress is possible. </p><p>　　In the literature on parametric modeling of productivity growth one can find decompositions comparable to the above (see e.g., Bauer [1990] or Nishimizu and Page [19

52、82]).</p><p>　　Figure 2. The input based Malmquist productivity index.</p><p>　　3. Results and comments</p><p>　　The data in this study consist of annual observations of outputs and

53、 inputs from 42 Swedish group pharmacies. The time period is 1980 to 1989. We specify four output variables and four input variables. Our four outputs: Drug deliveries to hospitals (SJHFANT); prescription drugs for outpa

54、tient care (RECFANT); medical appliances for the handicapped (FOLIANT); and over the counter goods (OTC). The first three outputs are measured in number of times. The volume of OTC is measured in 1980 prices. All</p&g

55、t;<p>　　Four separate inputs are used: Labour input for pharmacists (ARBTFT); labour input for technical staff (ARBTTT); building services (LOKY); and equipment services (AVSK). Labour input is measured in number

56、of hours worked. Absence from work due to sickness, holiday, education, etc., is excluded. The flow of building services is assumed to be proportional to the floor space available, measured in square meters. The services

57、 flow from equipment is assumed to be proportional to the stock of equipmen</p><p>　　The number of hours worked by technical staff has decreased by 32 percent on average between 1980 and 1989. On the other h

58、and, the number of hours worked by pharmacists is almost the same in 1989 as in 1980. However, on a year to year base, we observed small changes in the average number of hours worked by pharmacists on average. One reason

59、 for the decrease in hours worked by technical staff is that the pharmacies ceased to recruit new technical staff in the middle of the 1980’s. Another reason </p><p>　　Our estimation of services from equipme

60、nt shows an increase for the average pharmacy during the observed time period by 16 percent. We note that almost all pharmacies have changed from old to new equipment during the 1980’s and that some pharmacies have been

61、completely rebuilt. Looking at floor space we observe a decrease on average. One reason for the decrease in floor space is that the pharmacies have eliminated unnecessary space. Another reason is that new pharmacies are

62、smaller than the old </p><p>　　On average, our statistics show a decrease over time in deliveries of drugs to hospitals. A dramatic decrease of 57 percent occurred between 1983 and 1984, due mainly to reorga

63、nization.</p><p>　　Prescription of drugs for outpatient care has been fairly constant over time on average. However, we observe a peak in 1983, which is due in part to an increase in the out of pocket price

64、of prescription drugs that took place in December 1983. This increase led to an increase in sales of prescription drugs, i.e., prescription drugs that the patient should have purchased later but because of the increase i

65、n out of pocket price purchased earlier.</p><p>　　In the end of the 1970’s, medical appliances for the handicapped were introduced as new products for the pharmacy. On average an increase in sales of medical

66、 appliances took place during the 1980’s. Here, we note that the number of products for the handicapped has increased over time, which may be one reason for the observed growth in medical appliances for the average pharm

67、acy.</p><p>　　On average sales of over the counter goods have increased by 57 percent between 1980 and 1989 measured in fixed prices. An increase in over the counter goods is in accordance with the developme

68、nt policy pursued by these pharmacies during the 1980’s. The business of the pharmacies has gradually come to focus more and more on self medications. One may, however, find great variations among the pharmacies in the e

69、xtent to which this policy has been pursued.</p><p>　　The mix of inputs and outputs has changed quite considerably during our observation period. We also note that in all years we find differences in input m

70、ix and output mix across pharmacies.</p><p>　　Computer costs and expenditures for energy, cleaning, stationery, etc. are not included among our input variables, because the data is not readily available. In

71、addition, drugs and pharmaceuticals delivered to pharmacies have been excluded as inputs since the data is not readily available.</p><p>　　測度1980-1989間瑞典制藥業(yè)生產率變化：</p><p>　　基于非參數(shù)的Malmquist指數(shù)法<

72、/p><p>　　作者：R. FÄRE; S. GROSSKOPF; B. LINDGREN; P.ROOS</p><p><b>　　國籍：美國;瑞典</b></p><p>　　出處：生產力分析，1992年第3期，85-101頁</p><p><b>　　摘要</b></p>

73、;<p>　　在這篇文章中我們開發(fā)了一種基于Malmquist生產指數(shù)的非參數(shù)(線性規(guī)劃)的計算方法。探究該方法用于瑞典制藥業(yè)的情況。</p><p><b>　　1.簡介</b></p><p>　　本文的目的是為了開發(fā)一種非參數(shù)方法計算生產率增長的輸入模型，并將其應用到瑞典制藥業(yè)這個樣本中。我們的方法是由法瑞爾(1957)提出的效率測度和凱夫斯，克里

74、斯坦森和迪唯爾特(1982)提出的生產率的測量合并而成。在他的經典文章，“生產效率的測量”中，法瑞爾介紹了整體效率可以分解為兩部分組合效率的計量框架：資源配置效率和技術效率。技術效率是謝潑德(1953)和馬姆奎斯特(1953)的(輸入)距離函數(shù)的倒數(shù)，它是基于生產率指數(shù)的Malmquist輸入模型的關鍵構建模塊，我們在此處使用。凱夫斯，克里斯坦森和迪唯爾特(1982)定義的基于Malmquist生產率指數(shù)的投入比例，將其定義為輸入距離函

75、數(shù)。當他們把法瑞爾(1957)的整體效率和距離函數(shù)對數(shù)結構結合，他們展示了Törqvist指數(shù)可以來自兩個Malmquist指數(shù)的幾何平均數(shù)。在這里，沒有強加在行為或技術上的假設。相反，我們允許效率低下和分段線性模型技術。因此，我們的Malmquist生產力指數(shù)的區(qū)別在于生產前沿面效率的變化。這種區(qū)別證明對當前政策目的有效。</p><p>　　在瑞典，藥品零售業(yè)從1971以來便被公共壟斷。他們在與瑞典

76、政府達成的協(xié)議，Apoteksbolaget(瑞典全國性藥店)“是負責確保國內藥品供應充足。為此，企業(yè)應當利用此次契機發(fā)展自己，同時保持藥物的成本盡可能達到最低的水平”( 在1970年5月27日205號藥物零售貿易法案的第4條)。那就是，即要滿足需求又要使Apoteksbolaget成本最小化，這表明以輸入為基礎的生產力指數(shù)使用輸入距離函數(shù)是一個合適的方法。</p><p>　　目前，Apoteksbolaget

77、計算一個藥房的生產率為四個輸出變量的和兩類人員工作的總小時數(shù)的加權總和之間的比率。它分配的相應的輸出量被認為是反映了資源利用差異，并且在總產出的計算中所有的藥店都被分配相同的權重。通過總勞動投入的工作時間之和得到兩種類型的人員。對于每個藥店，生產力是由Apoteksbolaget在斯德哥爾摩的總部按月計算并且在三個月內反饋給藥店。藥店的報告包括與其自己一年前的生產力比較，以及與其他藥店的比較。目前，計算效率的方法對于分配給輸入、輸出值的

78、權值比較敏感，且它不會對所有輸入值賦權數(shù)。</p><p>　　我們的樣本包括42組(或區(qū)域)從1980到1989在瑞典經營的藥店。這些組中的藥房是在瑞典的藥店總數(shù)量的一小部分(在1989年瑞典有816個藥店)。我們選取這42組藥店有如下幾個原因。首先，我們有這42組藥店在整個時期的數(shù)據。第二，事實上他們都是屬于同一組群(相對于當?shù)厮幍昊蜥t(yī)院藥房)意味著他們的所負的責任和規(guī)模都非常相似。第三，這個樣本的適用性是用

79、由電腦計算而得的。</p><p>　　相對于目前Apoteksbolaget使用的方法，我們的方法測度生產率變化程度，以及在某些情況下改變測度方式有所不同。將樣本作為一個整體，我們發(fā)現(xiàn)生產率增長七期的同時生產力下降兩期。目前所采用的Apoteksbolaget使用方法也會使生產率前進七期同時倒退兩期。然而，最重要的一點是，時期的前進或倒退并不總是使用相同的方法到達的。例如，我們的方法表現(xiàn)出在1980和1981之

80、間平均是倒退的而在1985和1986之間平均是進步的。根據Apoteksbolaget的報告，這些年發(fā)生相反的情況。所以，一個藥店或一組藥店的平均值，在本文中提出的方法無論相對于水平或方向的變化，可能會對生產率變化得出截然不同的結果。</p><p>　　生產力和生產率的變化的計算方法目前大多采用Apoteksbolaget公司的方法，但有很多缺點，例如：(1)它假定潛在的制藥技術是一種特殊形式(并不是制藥生產技

81、術的一個適當?shù)募僭O)；(2)不能區(qū)分在效率和技術前沿面之間的變化；(3)它不能很容易地在同一單位中測量包括除勞動以外的多個輸入變量和需要輸出的變量；(4)它需要一個先驗的輸入和輸出變量集合的權重。所有這些缺點在使用我們的非參數(shù)方法的計算效率和生產率的變化中迎刃而解。</p><p><b>　　2.生產率指數(shù)</b></p><p>　　生產技術被定義在每個周期，，是

82、所有可行的輸入和輸出向量的集合。如果表示在周期的輸入向量，表示在相同周期的輸出向量，那么技術是集，其中。我們的技術模型通過輸入向量對應或等效的輸入需求集</p><p><b>　　(1)</b></p><p>　　輸入要求的集合表示在時期內所有的輸入向量都可產生輸出向量。這里我們假設是的閉凸集，但是是有條件的，例如。并且，我們可以對輸入向量和輸出向量分別進行處理，

83、例如和。</p><p>　　本文中，我們把方程(1)作為分段線性輸入需求集或等效地作為有效分析模型。在這個模型中的系數(shù)是由輸入向量和輸出向量的觀測值組成。我們假設在每一個時期，輸入向量，的觀測值有個。這些輸入向量是用于生成在時期，輸出向量，的觀測值，同時我們假設對于所有時期的觀測值都相同，例如。</p><p>　　要求輸入集(1)可從觀測值中形成</p><p>

84、;<b>　　(2)</b></p><p>　　其中是一個強度變量與有效分析中相似。強度變量是技術的形式，它是觀察到的輸入和輸出向量的凸錐。規(guī)模收益不變是基于基準技術，但其他形式的規(guī)模收益可以基于限制強度變量之和。也可以表明滿足以上介紹的屬性。</p><p>　　基于生產率指數(shù)的Malmquist輸入向量是由四個輸入向量的距離函數(shù)表示的。他定義為</p>

85、;<p><b>　　(3)</b></p><p>　　顯然，如下面的圖1所示。</p><p>　　在圖1中，輸入向量屬于輸入需求集。在可行的情況下，距離函數(shù)測量了的最大可能的收縮，例如。就圖1而言，。對于觀察的，，距離函數(shù)的值為線性規(guī)劃模型的解</p><p><b>　　(4)</b></p&g

86、t;<p>　　注意是一個元素的輸入集，這意味著距離函數(shù)的值大于或等于1。當輸入向量屬于輸入集合的等產量曲線時便實現(xiàn)了價值，因此它是技術上的有效α。</p><p>　　圖1. 輸入向量距離函數(shù)</p><p>　　我們注意到輸入距離函數(shù)是法瑞爾技術效率測度的倒數(shù)，所以我們利用距離函數(shù)計算技術效率。</p><p>　　為了確定輸入向量基于凱夫斯，克里

87、斯坦森和迪唯爾特(1982)提出的Malmquist生產率指數(shù)，我們需要在期的輸入輸出向量與后續(xù)時期的技術相關。因此，我們評價對于在期的輸入輸出向量的輸入距離函數(shù)與之后時期的輸入需求集相對應。</p><p><b>　　(5)</b></p><p>　　此處，。然而，向量不需要在期適用，因此如果方程(5)有一個解(例如，上確界最大)，的值會嚴格小于1。</p

88、><p>　　在我們的數(shù)據集中，輸入向量在每個時期的每個觀測值都是正值。這把處理后的投入向量與規(guī)模收益不變相結合以確保我們能計算, 的輸入距離函數(shù)(5)的值，作為線性規(guī)劃模型的解。</p><p><b>　　(6)</b></p><p>　　我們發(fā)現(xiàn)自從不必為輸入需求集中的一員，距離方程的值會嚴格小于1。</p><p>

89、;　　輸入距離函數(shù)需要兩個額外的評估為了確定生產率指數(shù)。我們需要評估t+1期的觀測值與t期和t+1期的技術相對應。特別是，</p><p><b>　　(7)</b></p><p><b>　　和</b></p><p><b>　　(8)</b></p><p>　　方程(

90、8)的計算與方程(3)相同，所以在方程(4)中我們只需要用代替。方程(7)的計算與方程(5)相似，所以我們也只需用代替，反之亦然。我們當然要注意，因為不需要在下可行，輸入距離函數(shù)會嚴格小于1。</p><p>　　接著凱夫斯，克里斯坦森和迪唯爾特(1982)的研究，我們定義基于Malmquist的生產率指數(shù)為</p><p><b>　　(9)</b></p&g

91、t;<p>　　實際上，我們的定義是凱夫斯，克里斯坦森和迪唯爾特(1982)定義的兩個Malmquist指數(shù)的幾何平均。</p><p>　　在他們的文章中，凱夫斯，克里斯坦森和迪唯爾特(1982)做了兩個假設。第一，他們假設和在每個時期的觀測值相等。在法瑞爾(1957)的術語中，這意味著沒有技術效率。第二，他們假設距離函數(shù)是相同二階對數(shù)的形式。這里我們遵循法爾特奧(1989)，且建模的技術為分段線

92、性并且允許低效率。因為允許低效率，生產力指標可以分解為兩個部分，一是測量效率的變化和其他是測量技術的變化或在前沿面上技術的等價變化。方程(9)可以改寫為</p><p><b>　　(10)</b></p><p>　　此處外部商值在技術效率低下改變時的等級測量和在期和期之間前沿面轉換時內部等級比率的測量如圖2所示。</p><p>　　我們用

93、表示期的技術，用表示期的技術，并且標注且有相同的定義。兩個觀察向量和在其各自的時期都是有效的。我們可以表示生產率指數(shù)在沿x軸的距離條件為</p><p><b>　　(11)</b></p><p>　　其中表示法瑞爾的技術效率測量比，最后一部分是和在技術變化的幾何平均。值得注意的是，技術的變化是在和的觀察值是局部測量得到的。這意味著：1)整體技術不需要表現(xiàn)一致，和2

94、)，技術退步是可能的。</p><p>　　文獻中生產率增長的參數(shù)模型可以找到一個分解與上述相媲美。</p><p>　　圖2.基于Malmquist生產率指數(shù)的輸入模型</p><p><b>　　3.結果和意見</b></p><p>　　這項研究中輸入輸出向量的數(shù)據是由42組瑞典藥店每年的觀測數(shù)據組成，時間跨度為1

95、980年到1989年。我們指定了四個輸入變量和四個輸出變量。四個輸出變量為：醫(yī)院交付的藥物(SJHFANT)；門診治療的處方藥(RECFANT)；殘疾人的醫(yī)療器械(FOLIANT)；和貨架上的藥物(OTC)。前三個輸出變量是以次數(shù)測度。OTC的容量是以1980的價格計算的。所有藥店對于一個給定產品輸出價格的變化都相同。</p><p>　　四個輸入向量分別是：藥劑師的勞動投入(ARBTFT)；技術人員的勞動投入(

96、ARBTTT)；房屋設備(LOKY)；和器械設備(AVSK)。勞動投入是以小時工作測度。由于生病，假期，和教育等因素導致的未能工作除外。房屋設備的流量被假設在地上的可用空間的比例用平方米來測度。器械設備的流量被假設為對設備的庫存比例。由假設，我們限制改變設備的庫存是正的或不變，除非一個藥店是完全重建。作為器械設備流量庫存的代表，我們使用1980價格測量制藥設備的年折舊。然而，因為我們只允許在庫存設備非負的變化，所以我們系列中每年的折舊以

97、不變價格計算測量，只顯示了增加或不變值。提出這個假設的主要理由是實際中多年使用的設備比會計記期的時間長很多。為了滿足會計的要求，我們假定設備只能提供八年的服務。</p><p>　　在1980到1989年之間技術人員的投入工作小時平均減少了32%。另一方面，藥劑師的勞動投入時間在1989年和1980年幾乎沒有什么變化。然而，與之前一年相比，我們發(fā)現(xiàn)藥劑師的平均工作小時平均有了小的變化。一方面的原因是技術人員的工作

98、時間減少，藥店在19世紀80年代中期便停止招募新的技術人員。另一方面的原因是一些技術人員已經培訓成為藥劑師。這種技術人員的培訓始于19世紀80年代初期。</p><p>　　我們對機械設備服務的估計顯示，在觀察時期內藥店的機械設備服務平均增加了16%。我們發(fā)現(xiàn)幾乎所有的藥店在19世紀80年代把舊的設備換成了新的，并且一些藥店已經完全重建了。就地上的可用空間來說，我們發(fā)現(xiàn)其平均在減少。一方面地上可用空間減少的原因是

99、藥店取消了不必要的空間。另一方面的原因是新的藥店比舊的要小。</p><p>　　平均來說，我們的數(shù)據顯示藥品交付醫(yī)院的時間減少了。主要由于重組的原因，在1983年至1984年期間大幅減少了57%。</p><p>　　門診的處方藥隨時間變化基本不變。然而，我們發(fā)現(xiàn)在1983年出現(xiàn)一個頂峰，這部分的主要原因是在1983年12月自費的處方藥價格上漲。這次上漲導致了處方藥銷量的增加，比如，病人

100、本應該在之后購買的處方藥由于自費費用的上漲便提前購買了。</p><p>　　在19世紀70年代末，殘疾人醫(yī)療器械作為醫(yī)藥的新產品。在19世紀80年代，醫(yī)療器械的銷量平均增長了。在這里，我們發(fā)現(xiàn)殘疾人的醫(yī)療器械產品的數(shù)量隨著時間的增加而增加，這可能是一般藥店的醫(yī)療器械數(shù)量增長的一個原因。</p><p>　　以1980年至1989年的不變價格計算，藥柜上的藥品銷量平均增加了57%。在19世

101、紀80年代藥柜上的藥品增加與藥店所推動的政策發(fā)展相對應。藥店里的業(yè)務逐漸越來越關注自我的藥物治療。然而，其中一個可能是政策已經實行對藥店有巨大的影響。</p><p>　　在我們的觀察期內輸入和輸出向量的組合發(fā)生了相當大的變化。我們也注意到全部年份中，我們發(fā)現(xiàn)輸入向量的組合與輸出向量的組合在藥店的差異。</p><p>　　能耗，清潔，辦公用品等的成本與支出的計算都不包括在我們的輸入變量之