Does the Internet Reduce Gender Gaps? The Case of Jordan

Abstract

This article investigates the link between digital technologies and female labour outcomes in a country with one of the lowest female labour force participation (LFP) rates. It exploits the massive roll-out of mobile broadband technology in Jordan between 2010 and 2016 to identify the effect of internet adoption on LFP, internet job search, employment and unemployment. Using panel data at the individual level and an instrumental variable strategy, the article finds that internet adoption increases female LFP and that the effect is driven by women who were not married in 2010, who also experience declines in marriage and fertility rates in response to internet adoption. An increase in online job search explains some – but not all – of the total increase in female LFP. Only women who are older and have higher levels of education experience an increase in employment in response to gaining internet access. The internet reduces the prevalence of traditional social norms among married women, but this channel does not explain the increase in female LFP.

Keywords:

• Internet
• Labour Market
• Gender
• Middle East

Acknowledgements

The authors would like to thank the participants at the MENA New Economy Authors’ workshop (July 2019, Washington DC), Jobs and Development Conference (June 2019, Washington DC) and WIDER Development Conference (September 2019, Bangkok) for valuable comments and suggestions. The findings, interpretations, and conclusions in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank Group, its Executive Directors, or the countries they represent.

Supplementary material

Supplementary Materials are available for this article which can be accessed via the online version of this journal available at https://doi.org/10.1080/00220388.2021.1965127

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. World Development Indicators (WDI), https://data.worldbank.org/indicator/SL.TLF.CACT.FE.ZS?locations=ZQ, accessed on 12 November 2020.

2. We also estimate the effects of increasing internet adoption on male labour outcomes to use them as a benchmark for the estimated impacts for women.

3. According to the JLMPS, more than 70 per cent of women needed to ask permission to go outside of the home in 2010.

4. In fact, the age at first marriage is 21 years on average, significantly lower than that of other developing countries (Winkler & Gonzalez, 2019).

5. Section 2 describes this literature.

6. Please see Krafft and Assaad (2018) for more information on these data.

7. We also exclude permanently disabled people from the data.

8. While population density could be an alternative control variable, official population density indicators are available at the governorate level only. We find, however, that the share of people living in urban areas in each governorate according to the JLMPS is positively correlated with density indicators at the same level.

9. Using an indicator of internet adoption or continuation at the individual level would be ideal, but the JLMPS only collects this information at the household level.

10. It is reasonable to assume that households that had fixed internet connections in 2010 had also adopted mobile broadband because such households were probably less affected by the barriers that typically restrict mobile broadband access, such as lack of affordability, education, and digital skills (GSMA, 2020; Prieger, 2015).

11. 28.3 per cent of observations in our sample gained internet access between 2010 and 2016, while 5.7 per cent reported having access to the internet both in 2010 and 2016.

12. Distance to a cell tower is linked with signal strength. The shape of that link is not obvious, however, since a variety of factors, including topography and environment, may affect it. We find that internet access was only weakly linked to linear distance to the nearest tower in the first-stage regression. For this reason, we express the distance to the nearest cell tower in logarithm, since this functional specification seems a stronger predictor of internet access in the first-stage regression.

13. Between 2010 and 2016, around 3 per cent of Jordan’s population continued to subscribe to fixed broadband internet (DSL technology).

14. The JPFHS interviews ever-married women aged 15–49 and has information on employment but not on LFP. We were able to construct a balanced panel of 60 subdistricts for the period under analysis.

15. Section 4 in the Supplementary Materials contains robustness tests of the main findings.

16. We define older women as 31–64 years of age. Because of the age restriction in our longitudinal sample (up to age 64 in 2016), adult women were 31–58 in 2010.

17. We choose high school as the cut-off in order to have samples with a more balanced number of observations and thereby to make sure that the differences between estimates are not driven by different sample sizes.

18. We did not find any impacts on male labour outcomes. See Table III in the Supplementary Materials.

19. Table 5 also shows that the LFP and employment of older women increased despite an absence of changes in their internet job searches. First, this group of women could have been more constrained by not being able to have a job that could be done remotely, which would be more compatible with household responsibilities, rather than by not being able to search for jobs online. Second, the first-stage regression in this subsample is weaker (with an F-stat of 8.37) which may lead to less precise estimates in the second stage.

20. We do not analyse impacts on wage levels due to the very low number of women reporting wages in the sample.

21. Section 5 in the Supplementary Materials describes the construction of social norms variables.

22. Results for the group of less-educated women should be interpreted with caution because the first-stage F-stat falls to 6.9 when restricting the sample to unmarried women in 2010.