What is the impact of mobile broadband on CO2 emissions from the international trade sector?

Climate change is one of the most pressing challenges of our time. Global temperatures are rising every year, causing both natural consequences and social threats. The World Trade Organization reports that international trade contributes to roughly 20 to 30 percent of total CO2 emissions. Because of globalization and increased trade, consumer preferences have changed while supply chains have become longer and more complex. To facilitate this level of international trade, larger amounts of fossil fuels and resources are used in production and logistic processes.

Figure 1 clearly shows that emissions in both exports and imports have increased during the last two decades. This underscores the critical need for innovations to mitigate this impact to reach global climate targets.

Among the various technologies examined for their potential to reduce emissions, information and communication technology (ICT) has received significant attention. In the ICT area, mobile broadband has changed the way we communicate and is today the most popular means of internet access. With mobile broadband being widely distributed and commonly used, it has the potential to reduce emissions through changes in production and consumption patterns. Increased adoption of mobile broadband technology can promote operational- and energy efficiencies, virtual communication, digitalization of processes, and smart logistics based on AI and IoT.

As seen in Figure 2, mobile broadband has an S-shaped technology diffusion curve, indicating that the number of users during the early diffusion stage is a relatively small proportion of all potential users. This study explores the role of mobile broadband in addressing climate change, analyzing whether its diffusion could lead to significant environmental benefits in the context of international trade. To my knowledge, there is a lack of consensus on the impact of ICT and mobile broadband on overall emissions, and there are currently no studies on how ICTs affect trade-related emissions. As part of my master’s degree in economics at Stockholm University, I have examined the effect of mobile broadband on CO2 emissions generated in countries' exports and imports, at Consumer & Industry Lab, Ericsson Research.

The study is based on a fixed effects framework using a data sample covering the 38 OECD countries from 2002 to 2018. Due to constraints in the data available on emissions related to trade, it has not been possible to include more recent years in this study, which has restricted the findings.

The results indicate that the relationship between mobile broadband and CO2 emissions embedded in imports and exports is significantly negative in a statistical sense. This indicates a significant reduction in emissions in trade at a country level as mobile broadband penetration increases.

Supporting regressions, which include control variables, suggest that the impact of emissions embedded in exports is slightly reduced. These regressions also indicate that the effect is likely to disappear one year after implementation for the median country.

Mechanisms through which mobile broadband influence trade-related emissions

In my thesis, I argue that mobile broadband can be categorized as a general-purpose technology (GPT), representing significant technologies that affect entire economies and have the potential to alter societies. As a GPT, mobile broadband could affect emissions through exports and imports in different ways:

• First-order effects refer to emissions that stem from mobile broadband life cycle and use, driven by energy-demanding investments in the diffusion process and infrastructure installation. First-order effects will likely increase emissions in both exports and imports.
• Second-order effects refer to productivity increases in different sectors due to technological diffusion. This means that fewer resources are required to generate the same output, which could potentially lower emissions in trade.
• Rebound effect refers to the change in consumer demand that stems from increased productivity. As productivity increases, the price of goods will probably decrease, resulting in an increased consumer demand. Depending on whether these goods are low-carbon alternatives or not, they could either decrease or increase emissions in trade. The reduction in prices gives consumers a rise in their disposable incomes, which could outweigh the trend to shop sustainably and increase overall consumption.
• Technology spillover: According to the New Economic Geography theory, firms are likely to cluster in urban areas where advanced technologies and infrastructure are available. This allows them to learn from each other and improve their production processes to emit less CO2. This could further decrease emissions from exports.

I consider that EU’s carbon market could give EU member countries within the OECD a comparative advantage in investing in technologies designed to reduce CO2 emissions. If mobile broadband can increase consumer demand for sustainable products, firms may adjust their production accordingly. The EU's carbon market, known as the EU ETS, sets limits on emissions for certain industries. Countries that exceed these limits may be penalized with a heavy fine. To avoid fines, firms in EU countries might increase their technology adoption to reduce emissions over non-EU countries.

Encouraging outcomes of mobile broadband adoption

Prior studies have examined how digital technologies affect emissions in certain industries, but none have focused on how ICT affects trade-related emissions. This thesis examines the relationship between mobile broadband and CO2 emissions in international imports and exports at the macro level, but lacks data to distinguish between first-, second- and rebound effects. Therefore, the examination of the factors influencing the impact of mobile broadband on emissions in trade has only been speculative.

The study is based on a two-stage model. I assume that mobile broadband diffusion is based on an S-shaped diffusion curve, similar to other types of technology adoption. This allows me to model the maximum penetration rate of mobile broadband by using the number of mobile phones and fixed internet subscriptions per 100 inhabitants in 2002, before mobile broadband was introduced in the economy, in a first-stage regression. Using the fitted values from the S-shaped diffusion curve in each country to estimate a causal effect from mobile broadband in a second-stage regression, I conclude that mobile broadband has a statistically significant negative effect on emissions in imports and exports. A 10-percentage point increase in mobile broadband penetration decreases emissions in exports by 10 percent and emissions in imports by 7 percent. It is important to note that technology networks negatively impact the environment, the ICT sector stood for 1,4 percent of global green house gas emissions in 2020. However, since this effect is quite small it is reasonable to assume that the emission-decreasing effects of mobile broadband use would outweigh the detrimental effects.

My findings are in line with previous research indicating that mobile broadband has an emissions-decreasing effect. The results show a statistically significant negative effect of mobile broadband penetration on emissions in exports and imports, with a stronger effect on emissions in exports than in imports. The decreasing effects could stem from second-order effects outweighing first-order-related emissions, as well as rebound effects negatively affecting countries' CO2 emissions. Mobile broadband could potentially promote energy- and resource efficiencies in energy-intensive sectors and alter the economic structure through changed consumption patterns. The stronger impact on exports might stem from a pattern in which firms using new technology tend to concentrate in specific regions where the technology is easily accessible. By doing so, firms can share their knowledge on using innovative technology to decrease CO2 emissions.

I perform supporting regressions controlling for GDP, population density, electricity mix, industry share, trade openness, a regulatory quality index, foreign direct investments, working age population, and human capital index. In one supporting regression, I included a time trend controlling for time-varying effects since the introduction of mobile broadband, suggesting that mobile broadband's impact on emissions in exports will decrease over time. This indicates that mobile broadband has diminishing marginal returns; once consumers learn to use the new technology, increased consumption and use will come to overshadow its second-order effects. This contrasts with findings showing a lagged effect, in line with past GPT studies, suggesting that energy-increasing rebound effects might offset emission reductions in the long term. With every new generation of mobile broadband demanding more energy, emission-reducing efficiencies and consumer demand for sustainable products becomes even more important. The supporting regressions are not as statistically significant as the second-stage findings, which may be due to the limited data sample.

Future research

The results imply that mobile broadband investments could be important in reducing emissions in international trade, however, more research is necessary to confirm this. As more recent data becomes available, future research should expand the analysis to include more countries and explore the effects of faster mobile broadband speeds offered by more advanced networks. Although the instrumental variable approach addresses endogeneity, further studies are needed to confirm mobile broadband's impact on trade-related emissions, including potential non-linear relationships and drivers behind the observed effects. 

Read more

Related research and policy:

How is mobile broadband intensity affecting CO2 emissions? – A macro analysis (ifn.se)

ICT’s potential to reduce greenhouse gas emissions in 2030

How 5G & connectivity can support climate action

More information on current IndustryLab research

Data sources:

OECD Data Explorer

GSMAi Sign in (gsmaintelligence.com)