(ProQuest: ... denotes formulae omitted.)1. INTRODUCTIONCo-existence of extreme economic affluence and significant poverty is main paradox of the modern world. This holds true for both across and within nation. It has given birth to a prominent theme of economic convergence in the existing literature. Convergence is defined as a long run tendency of poor economies to grow faster than the rich counterparts over time to equalize the per capita income (Solow, 1956; Swan, 1956). Not surprising, there emerged various theoretical developments on economic growth and convergence of economies over the years with varied explanations. Such as, Solow (1956) and Swan (1956) provide a fundamental base for convergence hypothesis, albeit indirectly. Mankiw et al. (1992), in their seminal work, offers an empirical validation of Solow's model and extended it by including human capital to investigate the convergence hypothesis. These deployments raises an important question that whether deconomies with lower capital labor ratio grows faster than the economies with higher capital labor ratio. Or in simple way, does there exists convergence across economies? (Barro and Sala-i-Martin, 2004).To answer this essential question there stands a plethora of literature using different version of the same. The issue has been debated in economic growth literature since long time and has been dealt empirically by numerous studies. Barro and Sala-i-Martin (1991), in their notable work tested convergence hypothesis and found evidences of convergence for the states of United States. Similar attempts were made in Barro and Martin (1995), Sala-i-Martin (1996), Martin and Sanz (2003).1 Later on the literature extended in using time series and panel data models (Crown and wheat, 1995; Bernard and Jones, 1996a, b; Carlno and Mills, 1993, 1994; Vohra, 1996) however, they lack unanimity among their findings. This provides an incentive to the researchers to adopt advanced and more appropriate approaches to deal with convergence hypothesis comprehensively for exploring new insights in this field.On the other hand there is a growing literature on examination and identification of the convergence clubs with the help of various methodologies.2 Studies like Quah (1993a, b; 1997), Friedman (1992) detected Galton's fallacy3 in convergence criteria used by earlier studies. Quah in a series of papers proposed an intra-distributional approach 4 to examine the dynamic convergence through twin peaks process. The method uses log differential of the interested variable in a stochastic kernels process to find out the clubs of converging regions in absolute as well as conditional to the other influencing variables. The empirical investigation of this approach by Carlno and Mills (1996) and Bernard and Durlauf (1995, 1996) supports the evidence of convergence hypothesis. There are some studies that made use of different methodology to identify club convergence, such as, regression tree approach (Durlauf and Johnson, 1995), pair wise approach (Pesaran, 2007), nonlinear log test approach (Philips and Sul, 2007) to name important ones and is still in debate.As far as India is concerned, its growth pattern has shown a significant variation over time. It lacks a primeval and very important concomitant of equitable disbursement over the states. Several attempts were made to understand the convergence behaviour of Indian states. Such as Cashin and Sahay (1996), tested the hypothesis of convergence on 20 Indian states between the period of 1961 and 1991 and found no significant signs of convergence. Similar results were reported by Bajpai and Sachs (1996). Whereas the studies of Marjit and Mitra (1996), Rao et al. (1999) and Aiyar (2001) found divergent behavior of Indian states. Further, Nagaraj et al. (2000) found evidence of conditional convergence for 17 states over the period of 1960-94. Trivedi (2002), Singh and Srinivasan (2002) found weak converging behavior across Indian states for 1990 to1998. …