بازده تولیدی تدارکات دولتی و خصوصی زباله های جامد و پیامدهای آن برای سیاست های مدیریت مواد زائد

This paper measures the productive efficiency of municipal solid waste (MSW) logistics by applying data envelopment analysis (DEA) to cross-sectional data of prefectures in Japan. Either through public operations or by outsourcing to private waste collection operators, prefectural governments possess the fundamental authority over waste processing operations in Japan. Therefore, we estimate a multi-input multi-output production efficiency at the prefectural level via DEA, employing several different model settings. Our data classify the MSW into household solid waste (HSW) and business solid waste (BSW) collected by both private and public operators as separate outputs, while the numbers of trucks and workers used by private and public operators are used as inputs. The results consistently show that geographical characteristics, such as the number of inhabited remote islands, are relatively more dominant factors for determining inefficiency. While the implication that a minimum efficient scale is not achieved in these small islands is in line with the literature suggesting that waste logistics has increasing returns at the municipal level, our results indicate that waste collection efficiency in Japan is well described by CRS technology at the prefectural level. The results also show that prefectures with higher private-sector participation, measured in terms of HSW collection, are more efficient, whereas a higher private–labor ratio negatively affects efficiency. We also provide evidence that prefectures with inefficient MSW logistics have a higher tendency of suffering from the illegal dumping of industrial waste.

History has shown that the inevitable consequence of economic growth is increasing waste. Waste, often dubbed as the “third pollution,” requires attention similar to air or water pollution.2 However, the waste processing industry, as well as households' decision-making processes regarding waste creation, appears to be subject to market distortions, such as hidden subsidies, ad-hoc regulations and inefficient public operations.3 When the market fails to discipline the industry, benchmarking and measuring efficiency of behaviors by decision-making units in the industry become important. In terms of cost, waste collection is the major component of municipal solid waste processing. For example, 74.7% of the total cost is due to waste collection in the metropolitan area of Tokyo.Nonetheless, it appears that not many economists have realized the importance of waste collection because the literature on this issue is very limited. Hirsh [7] was one of the first empirical studies to concentrate on the cost of waste collection, focusing on 25 cities near Saint Louis in the US. Following Hirsh's seminal work, Stevens [6] and Dubin and Navarro [13] studied larger samples in the US and showed empirical evidence of economies of scale [13] and economies of density [6]; Yamamoto [15] mostly confirmed similar cost structures for Japanese data.While these previous papers are based on parametric analysis, this paper measures the productive efficiency of solid waste logistics in Japan by applying various data envelopment analysis (DEA) models to cross-sectional data at the prefecture level. DEA, pioneered by Farrel [4], is one of the primary methodologies in estimating multi-output multi-input production efficiency. DEA nonparametrically identifies the production possibility frontier and then measures the inefficiency of each DMU as the distance to the frontier.5 A number of studies have adopted DEA in measuring efficiency for various industries, such as banking and transportation.6 However, the literature on efficiency measurements in the reverse-logistics industry is limited. The results consistently show that the set of underperforming prefectures is similar among different model settings, with Ehime being the most inefficient prefecture, closely followed by Nagasaki. Our list of underperforming prefectures indicates that small inhabited islands are overrepresented; therefore, geographical characteristics appear to be a major factor determining the efficiency of these underperforming prefectures. Ehime, for example, has 33 islands in its jurisdiction, with an average population of 525 and an average area of 2.71 km2 per island. This production size is too small to achieve the minimum efficient scale in waste collection because, as shown by the literature, increasing-return technology is implemented at the municipal level.7 At the same time, our results indicate that the production technology has constant returns to scale at the prefecture level. In addition to geographic influences, our results show that prefectures with a higher private sector participation, measured in terms of household solid waste collection, are more efficient, even though the labor ratio of the private sector negatively affects the efficiency. Through a spatial econometric analysis, we also provide evidence that prefectures with inefficient MSW logistics have a higher correlation with the volume of illegal dumping of industrial waste. This paper is organized as follows. Section 2 presents our data. Section 3 provides various DEA productive efficiency measurement results. In Section 4, we analyze the obtained results for the reverse-logistics industry in Japan. Finally, Section 5 concludes.

This paper measured the productive efficiency of the municipal solid waste (MSW) collection in Japan by applying DEA, the data envelopment analysis, to cross-sectional data of the fiscal year 2009 made available by the Ministry of the Environment of Japan. Our data include four outputs and four inputs for all 47 prefectures in Japan. The outputs are the volumes of household solid waste (HSW) and business solid waste (BSW) collected by both private and public operators, whereas the numbers of trucks and workers used by private and public operators are used as inputs. Either through public operations or by contracting or licensing private waste collection operators, prefectural governments possess the fundamental authority over waste-processing operations in Japan. Therefore, we estimate a multi-input multi-output production efficiency at the prefectural level via DEA, employing several different model settings. The results consistently show that the Ehime prefecture, followed immediately by Nagasaki, is the least-efficient prefecture, which indicates that geographical characteristics, such as the number of inhabited remote islands, are relatively more dominant factors of efficiency. While the implication that in these small islands' minimum efficient scale of production is not achieved is in line with literature suggesting that waste logistics has increasing returns at the municipal level, our results indicate that the production of waste collection in Japan is well described as CRS technology at the prefectural level. Through the regression of the measured efficiency scores of the public participation ratios, we have shown that prefectures with a higher private-sector participation measured in terms of HSW collection are more efficient. At the same time, the regression results indicate that the higher proportion of labor in the private sector implies lower productive efficiency. This result provides empirical evidence to advocate more private participation in reverse logistics, provided that the labor productivity of private operators is sufficiently high. We also provide evidence that the prefectures that are inefficient in MSW logistics tend to observe higher volumes of illegal dumping of industrial waste. Because the restoration of illegal dumping sites is costly, more investments in inefficient DMUs help minimize the cost of waste management policy.