As consumption expenditure or purchasing power is largely a function of household or individual income, wages are considered to be a robust indicator of the livelihood status of the population. Although, in a general neo-classical framework, wages are determined by the demand and supply of labour, developing economies are largely characterised by labour market dualism (Heckman and Hotz, 1986) and strong entry barriers across different segments of the labour market (Karan and Selvaraj, 2008). This dualism presumes the existence of two distinct sectors of economic activity, typically classified as the organised and unorganised sectors. While the organised sector generally offers more regular, stable and higher paid employment, the unorganised sector, on the other hand, is associated with irregular, unstable and low paid jobs.

In India, labour market dualism has been widely documented (Sen, 1994; Tendulkar, 2003; Das, 2003) with wages varying across different segments of the labour market (Sen, 1998). The explanations provided to account for the wage differential range from institutional factors, such as changes in government policy, to supply-side factors, such as demographic shifts, to demand-side factors, such as changes in the demand for skilled labour (Machin, 2002).

Using NSS data from the employment and unemployment surveys, we examine the wage structure for three categories, namely, rural and urban labour and salaried wage employment. The 61st (2004-05) and 64th (2007-08) NSS rounds coincide with the high growth phase, with India achieving a spectacular 9% growth rate.

On exploring the growth in wages across the three categories over the years, we find that wages of the lowest deciles in the rural labour category grew at a much faster pace than that of the top deciles. The same holds true for other categories as well. However, in most unorganised labour markets, wages are not indexed to the cost of living. As a result, a high rate of inflation acts as a tax, thus lowering the consumption expenditure (Datt and Ravallion, 1998a, 1998b). The authors have thus estimated the state-wise real wage growth rates for the above mentioned three categories.

In the rural labour category, real wages grew at the fastest pace in Andhra Pradesh, Karnataka, Kerala and Tamil Nadu, in line with the expansion of the farm sector (with the exception of Kerala). The relationship between real wages and poverty reduction is further underscored, as all these states also witnessed the sharpest reductions in rural poverty. On the other hand, states such as Rajasthan, Bihar and Chhattisgarh, which reported a marginal increase in real wages, witnessed lower poverty reductions in rural areas. In the urban labour category, real wages grew at the fastest pace in Jharkhand, Andhra Pradesh and Haryana, in line with the spectacular expansion of the non-farm sector. All these states reported sharp reductions in urban poverty. In Orissa and Rajasthan, a higher growth in real wages for urban labour translated into a larger decline in urban poverty estimates.

Although real rural wages grew at a faster pace in both MP and West Bengal, poverty reduction was more uniform only in West Bengal, while it was concentrated more in rural MP.

In the salaried employment category, Andhra Pradesh, Bihar, Tamil Nadu and Karnataka reported the largest increase in real wages. Jharkhand, in addition to witnessing a spectacular rise in urban labour wages, reported a rise of 7.8% in the salaried employment category. It is interesting to note that even with high growth in the non-farm sector, real wages actually fell in Delhi, Punjab and Uttaranchal.

Although, as mentioned above, wages of the lowest deciles grew at a faster pace than those of the top decile, in absolute terms wages received by individuals in the lowest deciles are significantly lower than the wages received by individuals in the top decile. There are several factors that are responsible for this, ranging from education attained, industry of occupation, location, etc. Given the sharp differences in wages, the authors have estimated wage inequality for the above mentioned categories. For each category, total inequality has been estimated using the general entropy class of inequality measures and then using states as the sub-groups, this total inequality has been decomposed into between-group and within-group, to study which states have witnessed a rise in the categories and which have not.

In the ?urban salaried employment? category, inequality at the all-India level has been more or less stagnant. On the other hand, in the labour category, inequality based on wages received by both rural and urban labour actually witnessed a decline. Interestingly, the fall in inequality in both rural and urban areas is roughly the same. However, this ?fall? conceals the heterogeneity in the direction of inequality in the categories at the state level. For example in the ?rural labour? category, inequality rose in Himachal Pradesh, Haryana, Bihar, Karnataka and Assam. In the ?urban labour? category, Himachal Pradesh, Punjab, Chandigarh, Delhi, Jharkhand, Orissa, Tamil Nadu and Karnataka reported sharp increases in inequality. In the ?urban salaried employment? category as well, inequality has sharply increased in Himachal Pradesh, Haryana, Uttar Pradesh, Bihar, Maharashtra and Karnataka.

While the period from 2004-05 to 2007-08 has seen a considerable reduction in poverty, it has also witnessed a widening of rural-urban per capita expenditure disparities. At the state level, consumption based inequality has, in fact, risen in quite a few states. Additionally, wage-based inequality has also revealed major inter-state imbalances, with inequality actually worsening in many states. While it can be argued that development does not start in every part of an economy at the same time and that it should be viewed as part of the growth process, the fact remains that it continues to widen even after two decades of reforms.

?Rajesh Shukla is Director, NCAER-CMCR. Ishan Bakshi is consultant, NCAER-CMCR