Profile‐based push models in manpower planning

Abstract

SUMMARYA methodology is presented to deal with heterogeneity due to observable variables in modeling personnel systems. For a manpower system, a model based on personnel profiles is introduced. The proposed algorithm concerns an analysis of the evolution of personnel profiles under time-discrete Markov assumptions. In this way, based on an historical personnel database, the estimation of transition probabilities of profiles as well as the computation of forecasts on the evolution of the manpower system can be found. Based on the presented procedure, a policy supporting software package has been developed for the Belgian Federal Government (research project financed by Federaal Wetenschapsbeleid).

Hidden heterogeneity in manpower systems: A Markov-switching model approach

“…In the considered manpower system, two grades (gr 1 and gr 2) are distinguished and no demotions occur. A study as in De Feyter (2006) or Guerry (2008) of the personnel dataset let us conclude that significant differences in transition probabilities are imposed by the observed variables gender and full time equivalent (FTE). More specifically, the outcomes female (f), male (m), FTE less than or equal to 0.5 (FTE  0.5) and FTE greater than 0.5 (FTE > 0.5) determine 8 personnel groups (concerning male employees) on the other hand form two subsystems for which in between there are no transitions.…”

“…In the first step of the procedure personnel groups k H H ,..., 1 homogeneous, with respect to transition probabilities in between the groups, will be determined by taking into account the available data on the observable variables. Those personnel groups k H H ,..., 1 can be found as the result of a multinomial logistic regression analysis (De Feyter, 2006) or by a profile based approach (Guerry, 2008). A person of the manpower system at time t is a member of one of these personnel groups )…”

“…In 2 previous work attention is paid to algorithms to detect the personnel groups within a manpower system that are homogeneous with respect to the transition probabilities (as there are promotion probabilities). So far those algorithms only take into account the information that is available from the personnel database on the observable variables (as there are grade, seniority, gender and full time equivalent) to determine the homogeneous personnel groups (De Feyter, 2006;Guerry, 2008). Nevertheless in a manpower system, apart from the observable sources of heterogeneity, there is heterogeneity due to latent sources (Ugwuowo and McClean, 2000).…”

“…The method will be formulated as a two-step procedure (section 3.1). In the first step of the procedure personnel groups, homogeneous with respect to transition probabilities, are determined by taking into account the available data on the observable sources (De Feyter, 2006;Guerry, 2008). In the second step heterogeneity caused by latent sources will be handled.…”

Evaluation of staffing policies in Markov manpower systems and their extension to organizations with outsource personnel

“…The books of Bartholomew (1982) and Vajda (1978) provide a comprehensive treatment of the theory of Markov models used in manpower planning, while the paper of Ernst et al (2004) provides a comprehensive review of the methods, models and applications in staff scheduling and rostering. A few of the contemporaneous and relevant papers are on the stochastic theory of compartments with applications to manpower planning (Agrafiotis, 1991), convergence of a manpower distribution (Ledermann, 1992), cohort analysis technique for long-term manpower planning (Chu and Lin, 1994), use of DEA for planning in UK universities (Sarricol and Dyson, 2000), modelling for planning and management of bed capacities in hospitals (Harper and Shahani, 2002), modelling heterogeneity in manpower systems (De Feyter, 2006), modelling of mixed push and pull promotion flows (De Feyter, 2007), an application of Markov manpower planning models in the armed forces (Skulj et al, 2008), profile-based push models (Guerry, 2008), a recent review of Markov models in manpower planning (De Feyter and Guerry, 2009a), Fuzzy set theory in stochastic manpower planning (De Feyter and Guerry, 2009b), two-stage workforce planning under demand fluctuations and uncertainty (Zhu and Sherali, 2009), and planning and benchmarking of doctoral programmes using Markov models (Nicholls, 2009). The most recent work is that of Li et al (2010) on optimal manpower decisions with a minimal length of employment constraint, Guerry (2011) on hidden heterogeneity in Markov manpower systems, Kim and Yoo (2012) on combined manpower planning and preventive maintenance strategies, and Guerry and De Feyter (2012) on optimal recruitment strategies in multi-level manpower planning.…”