Adoption of automation technologies

Adoption of automation technologies: Evidence from Denmark

Lene Kromann, Anders Sørensen 15 July 2019

The automation of production processes can be an important topic on the policy agenda in high-wage countries, but proof the economic ramifications of automation at the firm level is bound. This column presents insights on automation from new survey data for Denmark. The findings reveal that variation in the adoption of automation technologies is high, the change in adoption as time passes is slow, and almost half of Danish manufacturing firms relied greatly on manual production processes this year 2010. Increasing international competition from China is a driver for investments in automation.

Related

Increasing competition from low-wage countries has led many manufacturing firms to close or offshore elements of the production process. It’s been argued that may jeopardise continued welfare improvements. Helper et al. (2012) argue that the (US) manufacturing sector may be the major way to obtain commercial innovation and is in charge of the lion’s share of export earnings. The downsizing of the manufacturing sector is a cause for concern among policymakers in high-wage countries. They have already been searching for clever methods to recreate manufacturing production and jobs. New technologies and automation tend to be considered possible answers to the challenges. It really is expected that many technologies that can potentially donate to productivity will be developed later on (Council of Economic Advisers 2016). Additionally it is likely to benefit activities outside manufacturing through the ‘servitisation’ of products and closer connections to create and innovation, that allows for results for the full total economy (Bruegel 2017).

Existing focus on automation

Recent studies on automation have mainly centered on the relation between automation and employment. However, a few empirical papers study the partnership between industrial robots and performance include Graetz and Michaels (2018) and Kromann et al. (2019), and both papers find that the industry-level adoption of industrial robots has raised productivity.

There is nearly no systematic empirical evidence for the potential economic ramifications of automation at the firm level for modern times. You will find a small blast of existing literature on automation – Dunne (1994), Doms et al. (1997), and Bartelsman et al. (1998) – that targets a youthful wave of automation through the 1980s and early 1990s. These papers mainly describe differences between plants or firms that adopt automation technologies and the ones that usually do not. Bartel et al. (2007) study a far more recent period, namely, 1997-2002, but concentrate on one narrowly defined industry – US valve manufacturing – and discover that investments in automation enhance the efficiency of production.

New measures of automation

A deeper understanding requires firm data. For a fresh paper (Kromann and Sørensen 2019), we’ve gathered a fresh dataset that measures automation in Danish manufacturing firms. Thereby, we are able to present empirical evidence for the adoption of automation technology across manufacturing firms. Predicated on observations from firm visits, two important measures that describe automation stick out. They are the stock of automated capital and the share of production processes that’s automated. We label the latter gauge the automation score and the goal of this measure is to fully capture that automated capital may be used pretty much efficiently by the firms, based on how well it really is implemented and built-into the manufacturing system. The survey originated so that both of these aspects could possibly be measured. The motivation for like the automation score in the survey was that managers, suppliers of automation equipment, and skillfully developed claimed that the share of production processes that’s automated is an essential requirement of automation that’s not necessarily captured by standard measures of capital. This claim is strongly supported in the empirical analyses.

Findings

We begin by examining the adoption of automation across manufacturing firms. We find that the usage of automation is modest in lots of firms. In 2005, almost 40% of the firms’ investments in machinery and equipment targeted for automation were at approximately 10% or lower. This year 2010, almost half of the firms still relied to a higher extent on manual production processes. At the other end of the spectrum, there are firms that devote almost all their investments in machinery and equipment to automated capital and also have high automation scores. The email address details are observed in Figure 1. This leads us to summarize that there surely is high variation in the adoption of automation technologies across firms – a conclusion that also holds within sub-industries, various kinds of production, and firms with high and low export intensity.

Figure 1 New capital investments and automation in firms

a) Distribution of the percentage of new capital investments in machinery and equipment directed at automation

b) Distribution of the automation score for 2005 and 2010

Note: Predicated on 474 firms found in Section 4
Source: Authors’ survey on automation in manufacturing

Given the variation in automation across firms, an instantaneous question is what drives investments in automation. Therefore, we turn to examining a potential driver of the adoption of automation technologies. Specifically, we concentrate on increasing international competition from China since its admission to the WTO, and investigate whether it has accelerated the adoption of automation technologies. We make reference to this as the trade-induced automation hypothesis. We find that increasing Chinese exports to the world drive investments in automated capital, which supports the hypothesis. The firms that specialise in product types where Chinese exporters have a comparative advantage have a motivation to get more in automation to withstand the increasing competition in comparison to firms that specialise in other products. The growth rate in automated capital is 2.16% higher each year for the 75th percentile firm when compared to 25th ranked after Chinese export changes for the firms’ main product. We also investigate whether increasing Chinese exports to the world relates to the automation score. This works out never to be the case, suggesting that Chinese export changes for the firms’ main product only drive investments in automated capital however, not how these investments are implemented in production. The email address details are presented in Figure 2.

Figure 2 Automation and international competition: Five-year difference estimation, 2005-2010

Note: Figures are representations of Table II, columns 1 and 4 of Kromann and Sørensen (2019). Figures are constructed using the Stata programme “binscatter”. Each bin is represented by a dot and represents the mean of 26 firms. The amount of bins are chosen to satisfy Statistics Denmark’s regulations on anonymity that amongst others requires that both largest firms in a bin don’t have a complete sales share that exceed 85%. The fitted lines are estimated by OLS on the estimation data set comprising 442 firms. Regressions are performed on long-differences that sweep out firm fixed effects. The dependent variables are: Δlog of automated capital in the most notable panel and Δ automation score in underneath panel. The explanatory variables presented in the figure is Δlog of Chinese export within the merchandise type with the biggest sales share of the firm. In both figures, we control for a couple of explanatory variables including Δlog of other styles of capital, Δlog of employment, and Δ skill share and a complete group of industry by region dummies (10 industries and 8 regions). All changes are in five-year differences between 2005 and 2010, aside from Δlog of Chinese exports where in fact the five-year differences are measured between 2001 and 2006. The measures of international competition are measured at the merchandise level. There are 189 different product codes for the 442 firms.
Source: Authors’ survey on automation in manufacturing, UN Comtrade data, and register data from Statistics Denmark

Model description and external validation

The empirical model relates to the theoretical model suggested by Bloom et al. (2014) that proposes a positive relationship between innovation and import competition. Specifically, the theoretical model explains why firms that are more subjected to competition from China have larger incentive to innovate after trade is liberalised. The mechanism is driven by “temporarily trapped factors”, e.g. skilled workers that are costly to teach for the firm and, therefore, to fire because their firm-specific human capital is lost for the firm. The increasing competition from China lowers the demand for products that skilled workers produce. Because of the high training costs, the skilled workers keep their jobs in the firm, but their opportunity cost is reduced. Accordingly, the incentive for innovation in the firm increases after trade liberalisation as the opportunity cost of skilled workers has truly gone down. Bloom et al. (2016) establish empirical support for the model and investigate the result of Chinese import competition on innovation across twelve Europe.

Finally, we investigate if the automation measures are significantly connected with measures of firm performance. The analysis is ways to determine whether our survey contains important info on automation and not simply white noise. We find that increasing usage of automation is significantly connected with higher productivity growth and increases in profitability. The relationships are robust to an array of control variables including skill shares, other production factors, and industry dummies interacted with region dummies. These results offer some external validation of the automation survey.

Concluding remarks

As discussed above, the variation in automation across firms is high. A first-order policy question comes from this result, which is really as follows. Is the lack of adoption of technology sub-optimal and therefore motivates policy intervention or may be the lack of adoption optimal implying that firms usually do not automate because they’re specialised in products where production shouldn’t be automated?

The close collaboration with skillfully developed and production managers, during firm visits completed through the development of the automation survey, suggested that the reduced use of automation somewhat is due to a specific lack of the required skills and resources to research the firms’ needs, possibilities for automation, and automation planning the factory floor. The production managers weren’t unaware that automation technologies existed, however they were lacking knowledge or awareness regarding the precise technologies that they could spend money on, on how best to implement these, and which production processes to automate. In this sense, information barriers could be a significant market failure that potentially justifies policy intervention.

We find that increasing international competition from China is a driver of automated capital. Increasing international competition from China can, however, not explain changes in the automation score. A significant research issue for future years is drivers for the automation score answering the question why some firms have high automation scores while some have low scores.

References

Bartel, A, C Ichniowski, and K Shaw (2007), “How Does IT Affect Productivity? Plant-Level Comparisons of Product Innovation, Process Improvements, and Worker Skills”, Quarterly Journal of Economics, 122, 1721-58.

Bartelsman, E, G V Leeuwen, and H Nieuwenhuijsen (1998), “Adoption of Advanced Manufacturing Technology and Firm Performance in holland,” Economics of Innovation and New Technology, 6 (4), 291-312.

Bloom, N, P Romer, S Terry, and J Van Reenen (2014), “Trapped factors and China’s effect on global growth”, NBER Working Paper 19951.

Bloom, N, M Draca, and J Van Reenen (2016), “Trade Induced Technical Change: The Impact of Chinese Imports on Innovation, IT and Productivity”, Overview of Economic Studies, 83 (1), 87-117.

Bruegel (2017), “Remaking Europe: the brand new manufacturing as an engine for growth”, in R Veugelers (ed.), Blueprint Series 26.

Doms, M, T Dunne, and K R Troske (1997), “Workers, Wages, and Technology,” Quarterly Journal of Economics, 112 (1), 253-290.

Dunne, T (1994), “Plant Age and Technology use in U.S. Manufacturing Industries,” The RAND Journal of Economics, 25 (3), 488-499.

Graetz, G, and G Michaels (2018), “Robots at the job”, Overview of Economics and Statistics, 100 (5), 753-768.

Helper, S, T Krueger, and H Wial (2012), “How come Manufacturing Matter? Which Manufacturing Matters? AN INSURANCE PLAN Framework”, Metropolitan Policy Program, Brookings Institution.

Kromann, L, and A Sørensen (2019), “Automation, Performance and International Competition: Firm-level Comparisons of Process Innovation”, Economic Policy (Paper presented at the 69th Economic Policy Panel Meeting of Economic Policy in April 2019), forthcoming.

Kromann, L, N Malchow-Møller, J R Skaksen, and A Sørensen (2019), “Automation and Productivity – A Cross-country, Cross-industry Comparison”, Industrial and Corporate Change, forthcoming.

Leave a Reply

Your email address will not be published. Required fields are marked *