Process innovation in low-tech industries in India: An empirical exploration

Abstract

Till now, studies have determined the existence of process innovation using survey data; however, in this paper, we use secondary data to empirically establish its presence in low-tech industries. Our empirical approach consists of data envelopement analysis in the first step followed by regression analysis in the second step. We use a cross-sectional dataset, i.e., Annual Survey of Industries, 2017–2018 for our study. Our results suggest that among Indian low-tech firms, investment in computer equipment and software is the most popular strategy to trigger process innovation followed by investment in plant and machinery. We find that these factors have a greater impact on firms that export products and import inputs, than on firms that export products but do not import inputs. In other words, the export-import orientation of firms induces a differential impact of these factors on process innovation. One policy implication from the study is that government should encourage and incentivize low-tech firms to improve the capabilities of its labour force. This is the first paper in the Indian context that uses secondary data to empirically determine the existence of process innovation in low-tech industries.

Keywords:

• process innovation
• low-tech industry
• India
• computer and software equipment
• plant and machinery

Acknowledgements

I am thankful to the Editor in Chief and two anonymous referees for their useful comments, which have greatly benefited the paper. Needless to add, any errors that remain in the paper are mine.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 https://ec.europa.eu/eurostat/web/microdata/community-innovation-survey, last accessed on 18 March 2022.

2 The decision to focus on process innovation comes from the author’s work experience as a Chemical Engineer in the Indian Chemical industry.

3 This is inspired from the definition given by Reichstein and Salter (2006).

4 The National Industrial Classification (NIC) is an essential statistical standard for developing and maintaining comparable data base according to economic activities. The Central Statistical Organisation, Ministry of Statistics and Programme Implementation, Government of India publishes the NIC codes.

5 Diaz-Balteiro et al. (2006) note that if available data doesn’t include accurate measures about the outputs’ prices, then DEA input-oriented model can be chosen.

6 We have taken the factor variables in terms of share because of two reasons – we assume homogeneity with respect to technology and we want to empirically establish the presence of process innovation because of these factors at the industry level.

7 There is a huge literature on market structure and innovation (for e.g., Kamien and Schwartz 1982), which argues that market structure has an effect on innovation.

8 Equation 1 is a formulation with different intercepts and different slopes for each group. As a robustness check for our results, we also formulated an equation with same intercept and different slope for each group. The general nature of the results remained the same in both the specifications.

9 Since we ran separate DEA at each five-digit code level, we wanted to present the summary of these variables at the five-digit code level. However, presenting summary statistics for 226 five-digit codes will eat up a lot of space. Hence, in the interest of space, using the five-digit level data and results, the summary is presented at the industry level.

10 I thank the anonymous reviewer for pointing this out.

11 The general results remain the same for the same intercept and different slopes model. This is true for all the industries presented in this section.