Differences in returns to innovation by sector
In which sectors does innovation boost labour productivity most? Chapter 1 showed that product innovation is more prevalent in high-tech manufacturing sectors and knowledge-intensive services. However, these are not necessarily the sectors with the largest returns to innovation (see Chart 2.2).
On the contrary, returns to product innovation are particularly large for firms in low-tech manufacturing sectors (such as food products or textiles), where introducing a new product typically results in labour productivity more than doubling (for an example of an innovative firm in the food sector in Romania, see Case study 2.1). In medium-low-tech manufacturing sectors (such as plastic products and basic metals)15, introducing a new product is associated with a 126 per cent increase in labour productivity, while in high-tech and medium-high-tech (“higher-tech”) manufacturing sectors (such as machinery and equipment or chemicals) the average increase is 91 per cent.16
Read more
CASE STUDY 2.1.These effects are fairly sizeable, but they are not as large when placed in the context of the labour productivity distribution. A low-tech manufacturing firm with median labour productivity would move from the 50th to the 82nd percentile of the labour productivity distribution after introducing a new product. A higher-tech manufacturing firm, on the other hand, would move from the 50th to the 69th percentile of the labour productivity distribution.17
This variation in estimated returns to innovation can be explained by differences in the probability of introducing new products and the level of competitive pressures faced. Firms in high-tech manufacturing sectors are more likely to introduce new products (see Chart 2.3) and more likely to compete in national or international markets (as opposed to local markets). While these competitive pressures may explain why firms have greater incentives to introduce new products, they may also limit returns to innovation because such firms tend to be fairly productive in the first place. In low-tech manufacturing sectors, on the other hand, most innovations come from suppliers of equipment and materials,18 so low-tech firms’ ability to innovate depends crucially on their ability to adapt their production processes and the adaptability of their employees.19 The relatively small number of firms that manage to adapt and introduce new products successfully may manage to capture a larger market share as a result of their innovations, thereby increasing their output per worker. Some innovations by firms in low-tech manufacturing sectors may be due to firms moving production from China to eastern Europe owing to rising wage costs in China and the increasing cost of fossil fuels.20
Sam Mills
Sam Mills is an interesting case – an agribusiness company which has managed to significantly increase the value added by its products through substantial R&D activities.
Sam Mills is a Romanian group specialising in corn processing, corn-based food ingredients and, more recently, snacks and gluten-free products. The group’s first company was founded in 1994 and focused on corn milling. Sam Mills has grown over the years and now comprises a total of 10 companies with a wide range of activities, including the production and distribution of many different corn and pasta products.
Substantial investment in R&D activities since the mid-2000s has enabled the group to develop higher-value-added products such as feed, corn-based food ingredients and, more recently, healthy snacks and food products (mainly gluten-free pasta, cereals and products with a low glycaemic index). As a result, the group is one of the few companies in Romania that sells products through established retail chains in the United States, the EU and Asia (including chains such as Walmart, Wegmans and Delhaize), as well as selling products via Amazon and in specialist health food stores.
CHART 2.2
CHART 2.3
Source: BEEPS V, MENA ES and authors’ calculations.
Note: This chart reports the impact of innovation at firm level by sector, reflecting the impact on the dependent variable firm-level productivity, which is measured as turnover (in US dollars) per employee. The results are obtained by estimating a three-stage CDM model by ALS, where productivity is linked to innovation, and innovation, in turn, is related to investment in R&D. For a detailed description and the set of control variables included, refer to Box 2.1. The baseline model is adjusted slightly to account for the smaller sample size resulting from regressions on sector subsamples. State ownership variables are not included, as there are too few observations in some regions; the use of email is not included when explaining the incidence of R&D, as in some regions all firms make use of email. A robustness check on the baseline regression in Table 2.2 indicates that the main results remain valid after applying these adjustments. All coefficients associated with the impacts shown are statistically significant at the 1 per cent level. Sectors are based on ISIC Rev. 3.1. High-tech and medium-high-tech manufacturing sectors include chemicals (24), machinery and equipment (29), electrical and optical equipment (30-33) and transport equipment (34-35, excluding 35.1). Low-tech manufacturing sectors include food products, beverages and tobacco (15-16), textiles (17-18), leather (19), wood (20), paper, publishing and printing (21-22) and other manufacturing (36-37). Knowledge-intensive services include water and air transport (61-62), telecommunications (64) and real estate, renting and business activities (70-74).
Source: BEEPS V, MENA ES, Chart 2.2 and authors’ calculations.
Note: For definitions of sectors, see the note accompanying Chart 2.2.