A clarification of the use of multiple regression analysis in meeting the burden of proof in compensation discrimination litigation

Howard, Ryan Michael (Stellenbosch : Stellenbosch University, 2005-04)

Thesis (MComm)--University of Stellenbosch, 2005.

Thesis

ENGLISH ABSTRACT: The new set of employment equity laws call for South African organisations to justify their compensations systems. During compensation discrimination litigation, evidence is required to support arguments put before the court in order to meet the burden of proof. The similarity between foreign and domestic legal systems, suggests that the operational implications of foreign legislation will also be relevant to South Africa. This raises the debate as to the nature of fairness in the compensation context, the debate of comparable worth and the use of multiple regression analysis. The organisation must present to the court evidence to show that the choice of compensable constructs, their measurement and application does not discrimination directly or indirectly based on group membership. Multiple regression analysis, a statistical method to model the compensation system, is fraught with difficulties and misunderstanding. It is nevertheless the most appropriate method to investigate compensation fairness. Comparable worth and multiple regression analysis require assessment in the South African context. The issues, which hindered the successful use of multiple regression analysis abroad, are reviewed in order to smooth its entry into South African litigation. A framework is presented based on literature and case law whereby all parties concerned can produce and evaluate such evidence

AFRIKAANSE OPSOMMING: Die nuwe Anti-Diskrimineringswetgewing verlang van Suid-Afrikaanse organisasies om salarisstelsels te regverdig. Gedurende salarisdiskriminasielitigasie word bewys verlang om die bewyslas oor te dra. Die gelyksoortigheid van buitelandse en binnelandse regstelsels gee te kenne dat die operatiewe implikasies van buitelandse wetgewing relevant tot Suid-Afrika sal wees. Dit bevraagteken die aard van billikheid in die kompensasie konteks, die debat van vergelykbare waarde en die gebruik van veelvoudige regressieontleding. Die betrokke party moet bewys aan die hof toon om te bevestig dat die keuse van vergoedingskonstruksie, sowel as die meting en toepassing daarvan, nie onregverdig diskrimineer, ten opsigte van demografiese groepe me. Veelvoudige regressieontleding 'n statistiese metode wat gebriuk kan word om die salarissisteem voor te stel. Alhoewel dit vele onduidelikhede bevat, is dit steeds die mees toepaslike metode om salarisbillikheid te ondersoek. Vergelykbare waarde en meervoudige regressieontleding is in die Suid-Afrikaanse konteks geëvalueer. Die aspekte wat die sukses van die gebruik van meervoudige regressieontleding in ander lande verhinder het, is ondersoek en geëvalueer om die toekomstige toepassing daarvan in Suid-Afrika te vergemaklik. 'n Raamwerk gebaseer op literatuur en gevalle studies word voorgestel, waar al die betrokke partye sodanige bewys kan produseer en evalueer.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/50238
This item appears in the following collections: