TY - JOUR
T1 - Technical and scale efficiency in public and private Irish nursing homes – a bootstrap DEA approach
AU - Ni Luasa, Shiovan
AU - Dineen, Declan
AU - Zieba, Marta
N1 - Publisher Copyright:
© 2016, Springer Science+Business Media New York.
PY - 2018/9/1
Y1 - 2018/9/1
N2 - This article provides methodological and empirical insights into the estimation of technical efficiency in the nursing home sector. Focusing on long-stay care and using primary data, we examine technical and scale efficiency in 39 public and 73 private Irish nursing homes by applying an input-oriented data envelopment analysis (DEA). We employ robust bootstrap methods to validate our nonparametric DEA scores and to integrate the effects of potential determinants in estimating the efficiencies. Both the homogenous and two-stage double bootstrap procedures are used to obtain confidence intervals for the bias-corrected DEA scores. Importantly, the application of the double bootstrap approach affords true DEA technical efficiency scores after adjusting for the effects of ownership, size, case-mix, and other determinants such as location, and quality. Based on our DEA results for variable returns to scale technology, the average technical efficiency score is 62 %, and the mean scale efficiency is 88 %, with nearly all units operating on the increasing returns to scale part of the production frontier. Moreover, based on the double bootstrap results, Irish nursing homes are less technically efficient, and more scale efficient than the conventional DEA estimates suggest. Regarding the efficiency determinants, in terms of ownership, we find that private facilities are less efficient than the public units. Furthermore, the size of the nursing home has a positive effect, and this reinforces our finding that Irish homes produce at increasing returns to scale. Also, notably, we find that a tendency towards quality improvements can lead to poorer technical efficiency performance.
AB - This article provides methodological and empirical insights into the estimation of technical efficiency in the nursing home sector. Focusing on long-stay care and using primary data, we examine technical and scale efficiency in 39 public and 73 private Irish nursing homes by applying an input-oriented data envelopment analysis (DEA). We employ robust bootstrap methods to validate our nonparametric DEA scores and to integrate the effects of potential determinants in estimating the efficiencies. Both the homogenous and two-stage double bootstrap procedures are used to obtain confidence intervals for the bias-corrected DEA scores. Importantly, the application of the double bootstrap approach affords true DEA technical efficiency scores after adjusting for the effects of ownership, size, case-mix, and other determinants such as location, and quality. Based on our DEA results for variable returns to scale technology, the average technical efficiency score is 62 %, and the mean scale efficiency is 88 %, with nearly all units operating on the increasing returns to scale part of the production frontier. Moreover, based on the double bootstrap results, Irish nursing homes are less technically efficient, and more scale efficient than the conventional DEA estimates suggest. Regarding the efficiency determinants, in terms of ownership, we find that private facilities are less efficient than the public units. Furthermore, the size of the nursing home has a positive effect, and this reinforces our finding that Irish homes produce at increasing returns to scale. Also, notably, we find that a tendency towards quality improvements can lead to poorer technical efficiency performance.
KW - Bootstrapping
KW - D24
KW - DEA
KW - H51
KW - I19
KW - Ireland
KW - L33
KW - Long-term care
KW - Nursing homes
KW - Public versus private
KW - Technical efficiency
UR - http://www.scopus.com/inward/record.url?scp=84992731337&partnerID=8YFLogxK
U2 - 10.1007/s10729-016-9389-8
DO - 10.1007/s10729-016-9389-8
M3 - Article
C2 - 27787751
AN - SCOPUS:84992731337
SN - 1386-9620
VL - 21
SP - 326
EP - 347
JO - Health Care Management Science
JF - Health Care Management Science
IS - 3
ER -