The Life cycle Growth of Plants The Role of Productivity

The Life Cycle Growth Of Plants The Role Of Productivity-Free PDF

  • Date:28 Jun 2020
  • Views:3
  • Downloads:0
  • Pages:53
  • Size:582.83 KB

Share Pdf : The Life Cycle Growth Of Plants The Role Of Productivity

Download and Preview : The Life Cycle Growth Of Plants The Role Of Productivity

Report CopyRight/DMCA Form For : The Life Cycle Growth Of Plants The Role Of Productivity


accounting for sales growth volatility but their e ect is dampened by. wedges that are negatively correlated with these fundamentals De. mand is by far the dominating fundamental but technical e ciency is. non negligible especially at birth As plants age wedges have a less. dampening e ect and demand di erences increasingly dominate sales. The dominance of demand is driven by superstar plants that are in. the top quartile of life cycle growth In contrast the lowest quarttile. of plants are driven by weak technical e ciency Lumping wedges. and costs inclusive of technical e ciency yields the misleading view. that costs are not important, Keywords post entry growth TFPQ demand distortions. JEL codes O47 L11 O14 O39,1 Introduction, A prevalent feature of market economies is wide heterogeneity of rm size. rm growth and a host of rm attributes correlated with size e g pro. ductivity exports and survival What are the sources of rm size and. rm growth heterogeneity Motivated by wide di erences in the distribu. tions of rm size and rm growth across levels of economic development the. macro literature on misallocation studies the role of productivity vs resid. ual wedges with special focus on wedges in driving size di erentials e g. Hsieh and Klenow 2009 2014 HK henceforth 1 Other literatures in macro. trade and IO have focused on the role of di erent attributes of rms de. mand quality markups and costs frequently analyzing them separately 2. Hottman Redding and Weinstein 2016 henceforth HRW recently devel. oped and estimated a framework where demand markups and residual costs. are simultaneously accounted for nding a dominant role for demand at. Restuccia and Rogerson 2008 and Hsieh and Klenow 2009 are seminal contribu. tions but that literature is extensive Examples are Guner Ventura and Xu 2008 Midri. gan and Xu 2013 Bartelsman et al 2013 Bento and Restuccia 2017 Adamopoulos. and Restuccia 2014 Eslava et al 2013, Quality is the focus in Brooks 2006 Fieler Eslava and Xu 2018 Hallak and Schott. 2011 Khandelwal 2011 Kugler and Verhoogen 2012 Manova and Zhang 2012. Production e ciency is emphasized in many of the applications of the Melitz model 2003. e g Eslava et al 2013 Markups have been emphasized by De Loecker and Warzynski. 2012 and De Loecker et al 2016, tributes 3 Wedges i e deviations from the model are not considered in. HRW s accounting framework, Building on these distinct approaches we develop a conceptual mea.
surement and estimation structure that takes advantage of uniquely rich. data to measure not only idiosyncratic demand shifters quality appeal and. markups but also two distinct dimensions of idiosyncratic marginal costs. technical e ciency and input prices Our framework accounts for the con. tribution of each of these attributes of rms to rm size and growth while. also allowing for wedges between the data and the behavior predicted by the. model We thus e ectively decompose growth over a plant s life cycle into. that attributable to shocks from demand technical e ciency input prices. idiosyncratic markups and residual wedges We apply this framework to. the analysis of growth over the life cycle of manufacturing plants in Colom. bia Life cycle business growth is crucially related to aggregate productivity. growth HK 2014 and displays wide heterogeneity across businesses Halti. wanger Jarmin and Miranda 2013, Crucial to our approach is detailed data on quantities and prices for out. puts and inputs which we obtain from the Colombian Annual Manufacturing. Survey This is a census of non micro Colombian manufacturing plants with. data on quantities and prices at the detailed product class for outputs and. inputs within plants Individual plants can be followed for up to thirty years. 1982 2012 The availability of price and quantity data for both outputs and. inputs at the product level permits separate measurement of fundamental at. tributes of plants on the technology the demand and the cost sides as well. as idiosyncratic markups The long time coverage allows us to investigate. the determinants of medium and long term life cycle growth. By technology or technical e ciency we refer to a production function. residual where production in multiproduct plants is plant level revenue de. ated with a quality adjusted plant level de ator We will refer to this tech. nical e ciency dimension as T F P Q as in Foster Haltiwanger and Syver. son 2008 though we generalize the concept to producers of heterogeneous. goods 4 On the demand side we estimate plant speci c demand function. Foster et al 2008 2016 also integrate demand and e ciency shocks in explaining. rm performance as does Gervais 2015 in the context of explaining rm exports A. prominent role for demand is found though not as dominant as in HRW perhaps as a. consequence of the direct measurement of e ciency, Hsieh and Klenow 2009 2014 use the term T F P Q to refer to a composite productiv. ity measure that lumps together technical e ciency and demand shocks We refer to this. residuals that identify greater appeal quality as the ability to charge higher. prices per unit of a product HRW Khandelwal 2011 Fieler Eslava and Xu. 2018 Input costs are directly measured from input price data Our spec. i cation of demand and competition allows for idiosyncratic markups that. vary with the plant s market share and with the elasticity of substitution in. the plant s sector, Our approach requires and the richness of the data permits estimat. ing the parameters of the production and demand functions for each sector. both to obtain T F P Q and demand appeal as residuals of these functions. We introduce an estimation technique that jointly estimates the production. factor elasticities and the elasticity of demand bringing together insights. from recent literature on estimating production functions using output and. input use data and literature on estimating demand functions using P and. Q data 5 As in the former we use a proxy approach to form moments that. identify production function coe cients 6 As in the latter we rely on supply. shocks to identify the slope of the demand function But in contrast to much. of that literature we identify the slope of the demand function by assuming. that current period innovations to technology are orthogonal to lagged de. mand shifters in levels We thus allow T F P Q and demand to be correlated. even over time This would be the case if as plausible improving quality re. quires greater e ort in the production side or investments in improving plant. attributes depend on previous pro tability Estimating production and de. mand jointly ensures consistency and thus proper separate identi cation of. revenue vs production parameters Moreover the granularity of our data. allows estimating di erent production and demand elasticities for di erent. sectors and without imposing constant returns to scale. composite concept further below as T F P Q HK as a reference to Hsieh and Klenow. Haltiwanger Kulick and Syverson 2018 explore properties of T F P Q HK using U S. For production function estimation see e g Ackerberg Caves and Frazer 2015 De. Loecker et al 2016 For demand function estimation see e g Hottman Redding and. Weinstein 2016 Foster Haltiwanger and Syverson 2008. Our approach relies on the assumptions that permit estimating the joint produc. tion demand system without specifying the nature of any wedges that impact the evolution. of the size distribution This is common in the literature although there are some excep. tions Cooper and Haltiwanger 2006 for example consider a speci cation of adjustment. costs that yield a multiplicative disruption e ect on productivity and pro tablity We. are implicitly considering separable non internal adjustment costs that would manifest. themselves in wedges, Manufacturing plants typically produce multiple products and use mul. tiple material inputs Moreover there is ongoing product turnover in both. outputs and inputs De ning and measuring real output and inputs at the. plant level thus requires constructing plant level price indices for both out. puts and inputs We follow the insights of Hottman Redding and Weinstein. 2016 and Redding and Weinsten 2018 and rely on a nested demand. structure at the product level within plants to build plant level price indices. that allow for turnover and shifting appeal across products and inputs within. After estimating plant speci c technical e ciency demand product ap. peal shifters markups and input prices we measure the contribution of each. to the variability of sales growth across plants over the life cycle Residual. wedges in our framework correspond to the gap between actual size at any. point of the life cycle and size implied by the di erent fundamentals 7 Since. we explicitly account for idiosyncratic input price and markup variability the. distribution of these wedges is not captured by revenue product dispersion. in contrast to Hsieh and Klenow s 2009 2014, Post entry growth is found to be highly dispersed and skewed in our data.
as it is in other contexts e g Decker et al 2014 2016 By age 20 plants. in the top quartile of predicted revenue growth have multiplied their sales by. a factor of 4 9 relative to their birth while those in the lowest quartile grow. by a factor of 1 37 Our focus is on decomposing the substantial variance in. growth across plants at di erent stages of the life cycle. While revenue growth is widely disperse the variance of growth of mea. sured plant attributes is far greater As a result residual wedges are strongly. size correlated while superstar plants actually grow less than implied by their. appeal and e ciency growth plants in the lowest two quartiles of fundamen. tals growth display positive wedges between actual and predicted growth. Correlated wedges are particularly large for this last group The top quartile. of predicted growth would have grown close to seven fold in the absence of. wedges by age 20 while revenue for the bottom quartile of predicted revenue. growth would have contracted markedly compared to birth Negatively corre. lated wedges reduce plant revenue variance by 15 percentage points over the. rst twenty years of life While residual wedges are far from negligible this. These wedges are also frequently termed distortions but we prefer the former term. since the idiosyncratic gaps we identify may represent sources of productivity or welfare. loss that even the social planner would incur as they may stem from constraints more. technological in nature such as adjustment costs, magnitude also implies that growth of fundamentals is still the dominating. factor in explaining the distribution of revenue growth. Rapidly growing demand shifters are the key di erentiating attribute of. superstar plants the third quartile of predicted growth exhibits appeal demand. growth that is less than half of the top quartile while plants in the lowest. two quartiles barely exhibit appeal growth Though our nding of a domi. nant role of demand shocks in accounting for life cycle growth in Colombia is. consistent with previous ndings for the US HRW and Foster Haltiwanger. and Syverson full distribution accounting allows us to identify this role as. stemming from extremely dynamic appeal in superstar plants Moreover our. results also point to a technical e ciency e orts as a necessary condition for. success rapidy contracting T F P Q is the outstanding characteristic of worst. performers in the bottom two quartiles, We also nd that these patterns vary considerably over the life cycle. For mature plants most of the variation in life cycle growth is explained by. measured fundamentals while for younger plants TFPQ and wedges nega. tively correlated with fundamentals play a more important role in the de. composition of variance The diminished role of TFPQ for more mature. plants partly re ects an increasing inverse correlation between TFPQ and. demand product appeal over the life cycle That is though superstar plants. are those with very high product demand appeal producing such products. is associated with reductions in TFPQ, We contribute to the literature in di erent ways First we bridge the gap. between distinct approaches to the study of drivers of rm size and growth. alternatively focusing on wedges with respect to productivity broadly de ned. to encompass both e ciency and demand or on the roles of demand cost. and markups Our framework builds on HK on the supply side and allows. for wedges a la HK and builds on HRW on the demand side Our ndings. yield insights that are masked by taking the two approaches independently. Our results imply that cost factors play a more important role than would be. identi ed by the HRW approach because their cost component is ultimately a. residual that lumps together e ciency input costs and all other unmeasured. factors negatively correlated with the former in our context Relative to the. implications of the HK decomposition our approach yields insights masked. by using a composite productivity measure As highlighted above TFPQ. and demand have very di erent contributions over the life cycle and across the. distribution of growth rates Our approach also enables further re nement. of residual wedges by breaking out the contributions of input prices and. Within the misallocation literature recent contributions have increasingly. focused on decomposing size to productivity wedges into components such as. The Life cycle Growth of Plants The Role of Productivity Demand and Wedges Marcela Eslavayand John Haltiwangerz June 28 2019 Abstract What determines rm growth We develop and estimate a frame work that examines the roles of technical e ciency input prices de mandshocks idiosyncraticmarkups andwedges Previousapproaches havecombinedtechnicale ciencyandproductappeal

Related Books