Productivity modeling in the apparel industry of Bangladesh: an application of data envelopment analysis (DEA) technique

Abstract

Although labor prOdU~li\ily Or capital producLivity are popular 111eaSUl'eSfur ilny production sy:<km, it portrays only" parlial picture 0 r iIs performance, Comparison of performances of producLioll ~ystcm<; is a nlUst for bl.lh<,cqucnt improvement effort. l-Ilmevcr, perfllrmancC measurement, ~,>pcciall},In complex production environment. has remained a much sought about topic among re,e<lr~her<;over the de~ade.\_ Several indices hm'e been proposed under \aricd Ci[~llmstances, but none gained unanimous acceptability. Several modds have ~lso han propo,ed under different condition, and e~vir()nmcnt, Data Envelopment Analysis (DEA) is a non parametric analyti~;ll model. which Cilll meawrr the perJemnances of production, as wdl as service ,ystem" taking into account multiple variables Researches re\eul that DEA 1l;,s bcen applied by researchers in different type~ of production systems, As reports reveal, although ratio analysis ofplOductivity has been applied to anal}ze producti\'it} and its \\(eak linkages in textile industry of Bungladesh, the DEA, a much betler anal}tical ICchnique, has not been applied to anal}~e thc performance~ of any production system. The appard factorie~ in Bangladesh suffer from poor prodLlctivity due to several factors. or variables. On many occasions. these variables are not only complex in natme by itself, but intcracting too, ther~by multiplying the complcxit}, further, fhese factors or \'ariablcs have uever been analyzed econometrically, As a result. accurate performance. in terrn.1 of producti\'ity. could ne,'er be known, This imp~des sub,e'luent improvement drive. This research aimed at analyzing th" performances, finding out the weak variable linkages and identifying the efficient frontier of apparel industry of Bangladesh. Utilization of input quantity and d"ticiency of the production system to maximize output need serious consideration. Nevcrtheles>, thi, has not gained due attention from the rebeurchers, In thi, research, the input and output oriented models for both COn.ltantretul'lls to scale a<,well as variable retmns to scak have been a~alyzed to find the relative scores of the productive efficiency of several apparel factories. From the scores of efficiency measurement, the most cfficient production periods (months) have bcen obtained. The rest of tile inefficient pcriods have the scope to elevate their bcores eithcr by decreasing their input or by increasing their output, in order to bceome productive periods. Thc next step of analysis has been.performed by applying the sluck based model. In this model, both the input and OlllPLltquantitie, have been dealt with simultaneousl}, i,e. to decrease the input and at the same time to increase the output, This type of analysis is expected to providc results better in the sense that unlike the prcvious model, both the input and output treatments have been pO'isiblc simultaneously, The sensitivity analysis indicates the efficiency zones between which the firms can be operated without losing their productive efficiency values. In other -viward" the ma:dmum possible conlraetian or e:>.pansioll of the input or aUlpllt quantities nwy h~~ollle pos~ihle within thi.\ safe range. Scnle efficiency is an import,lllt parameter to Judge from it> value when under unity, there exists a ,cope lo ill~r"nse quantity of production. Malmquist Prodl\ctivity lndex with gr~atcr value of lIllity sh,m's that there i~ grov,1h in productivity comp"red [0 it, earlier period. In this r",search. filleen parameters havc been considered in order to detcrmine their influences upon the outPllt of the workers as a whole. Five among them have bccn round to be inl1llcncing the outpllt produced, which arc: Gcnder. Age Group, Work Experien~es, Satisfaction~ and Qualifi~aliorls ofthe workers.