Innovation and development: a cross-country comparison

In October 2013, Jun Hou successfully defended his PhD thesis on ‘Complementarity in Innovation and Development: A Cross-country Comparison’. Focusing on several thousand firms in more than 30 developing countries, he provides a range of recommendations linked to commonly found obstacles: competition, finance, skilled labour and tax regimes. In answering the questions below, Hou explains his research and how he made use of massive datasets from the World Bank.

1. You mention the “synergistic effects of innovation strategies and policies in developing countries”. What countries did you look at? Did they apply different strategies and policies to similar problems? What were the results?

JH: The empirical analyses of the dissertation mainly focus on technological sourcing strategies by using firm-level data from the Chinese manufacturing sector and in 24 other developing countries. The corresponding innovation policies for innovation obstacles are investigated later in the dissertation, where sample firms are drawn from 31 developing countries in the regions of Eastern Europe and Central Asia. Despite innovation policy schemes having to be tailored to countries’ and firms’ specifications, our findings highlight the critical role of external technological sourcing to firms’ innovation gains across developing countries.

2. One of your data sources was the Investment Climate Survey from the World Bank. How many countries and sectors does this cover? How reliable are the data and the survey overall?

JH: the Investment Climate Survey (ICS) was collected by the World Bank in 135 countries from different regions of the world between 2002 and 2011. The data cover major two-digit manufacturing industries according to the International Standard Industrial Classification (ISIC), revision 3.1. The surveys are divided into several waves by the year in which they are conducted. According to the standardized questionnaire, two broad waves are defined: 2002-2005 and 2006-2011. The wave of 2002-2005 covers 63 countries; the wave of 2006-2011 included 114 countries. For each wave, there is a standardized questionnaire that covers a set of questions for firms in both manufacturing and service sectors that are addressed in each country’s survey. In general, questionnaires used in the survey are the same, or at least very similar, within a region. Larger differences are detected across regions due to methodological and survey instrument changes across regions and years.

The ICS features questions designed in the context of developing countries. It considers the fact that innovation in developing or least developed countries is a catch-up process through learning, adoption and imitation. Great attention is placed on learning activities whereby firms have acquired technology, particularly in the wave of year 2002. This unique information has assisted us to conduct an insightful analysis with respect to technological–sourcing behaviour of firms in developing countries. However, a major shortcoming is that a large number of inconsistencies appear such as currency inconsistencies, missing values or unreasonable answers, even though there are always improvements of the quality and accuracy of the data. Hence, thorough familiarity with the questionnaires and careful scrutiny of the data are essential before proceeding with further analyses.

3. Among Chinese manufacturing firms, you found that in-house R&D accelerates the assimilation of external know-how, while external technologies improve the efficiency of in-house innovation. Which is more important to their viability and sustainability? Or is the relationship truly symbiotic?

JH: By studying Chinese SME manufacturing firms, we found that firms with R&D are more innovative, and the same applies to firms with technology purchasing. It is always difficult to draw a concrete conclusion on which technological-sourcing is more viable and sustainable given the fact that these behaviours are not only directly fostering innovation output, but also accumulating valuable experiences, which consequently upgrade firms’ technological capability. Here, instead of trying to capture a comprehensive effect caused by conducting in-house R&D and/or technology purchasing, our interpretations are based on the estimated marginal effects of the innovation function. It suggests that in-house R&D contributed more than technology purchasing to the propensity of a firm to become an innovator.  With limited resources to allocate to innovation, firms in developing countries (especially small-sized firms) are unable to participate in both R&D and technology purchasing, even though both are expected to exhibit positive effects on innovation gains. While firms have become bigger and established adequate absorptive capacities, synergetic effects may appear.

4. You tested the potential synergistic effects between technological strategies by using firm cross-sectional data from 24 developing and least developed countries. What countries were most impressive in this respect? What were they key synergies identified?

JH: With the Investment Climate Survey data, we took a close look at the internal and external technological-sourcing strategies in 24 developing countries. Borrowing the Meravete and Pernias (2007) method, this chapter emphasized the separate identification of synergistic effect and the unexplained correlations between the error terms. The findings confirmed the potential productivity gains of participating simultaneously in internal and external innovation activities, and that countries in the middle-income group are more impressive in this respect. However, this synergistic effect is subject to the technological capability of local firms. By dividing the sample into two groups according to their income levels, we found that firms in low-income countries rely more on external sourcing, while firms in middle-income countries rely more on in-house R&D. It is reasonable to speculate that low levels of technological capability severely constrain firms from low-income countries in their efforts to build up their own R&D stock of knowledge, instead making them rely on external knowledge acquisitions such as licensing, hiring skilled labour, or purchasing technology. Hence, at a low level of development, governments should instigate policies that encourage external sourcing, such as providing technological information and financial support for technology imports. In middle-income countries, where firms have sufficiently built up their innovation capabilities and gathered adequate experience, government policies should focus on encouraging firms to invest in both internal and external innovation activities. This finding is consistent with conclusions drawn on Chinese manufacturing firms, which suggest that prior accumulated innovation experiences may be necessary for firms to utilize external know-how.

5. What barriers to innovation did you find? What policies or combination of policies can help to overcome this in developing countries? Any notable examples?

JH: There are four kinds of obstacles to innovation — finance, skilled labour, competition and tax — investigated in around 7000 firms from 31 Eastern European and Central Asian countries. Manufacturing firms, companies in the higher-income group, and the larger-sized group are more likely to succeed in innovation and less likely to be burdened by obstacles. Meanwhile, low and low-middle income groups face pressures linked to inadequate finance: more than 17 per cent of firms perceive it as a severe obstacle. This is respectively about 5 and 17 percentage points higher compared to upper middle-income and high-income groups.

The potential interactive effects among obstacles are notable from a policy perspective. Therefore we take a step further to estimate the potential complementarities among obstacles. If there exists complementarity among obstacles, a policy that focuses only on one obstacle will be assumed to be insufficient. A policy package will be more effective and necessary in this context.

The findings show that four kinds of obstacles are complementary with respect to the propensity of innovation. This suggests that additional obstacles are irrelevant if existing obstacles already prevent a firm from becoming a product innovator. Hence, policy instruments should be designed to eliminate obstacles together rather than removing them individually. In contrast, a substitutive relation – characterized by finance and tax related obstacles – plays a dominant role with respect to the innovation intensity. Under these circumstances, the presence of one obstacle (e.g. high tax rates) exacerbates the undermining effects induced by the other obstacle (e.g. inadequate internal finance). Policies that seek to eliminate them individually are therefore recommended. Click here for more photos of the PhD defence ceremony.

1383076_377445185691440_1216971162_nDr. Jun Hou obtained his MA in Economics of Development from the Institute of Social Studies. He previously interned with the Global Programme of Action office at the United Nations Environmental Programme. He is now appointed as a research officer at TMD centre, Department of International Development, University of Oxford.