Social insurance: global, exclusive, overstated as an institution

The introduction of work-injury insurance is seen by many welfare state scholars as a pivotal moment in each country’s history. It created two new institutional features of societies.

1. Work-related risks were re-framed as affecting all of society rather than individual, familial or employer-specific risks. They were redefined as social risks – those that inevitably face all of society across time and space.

2. The state took on the role of arbiter of social risks, by developing, mandating and regulating risk-pooling in the form of social insurance – nationally mandated, regulated and/or provided insurance. A process that usually started with regulating work-related accidents.

“Accidents no longer seemed an interpersonal matter to be sorted out between workers and employers in court. Instead, they became a social problem and a target for social policy.”

(Moses 2018:4)

After its late 18th Century origins in Europe, work-injury laws in the form of social insurance were written into the law books of nearly every country in the world.

Fig 1. The introduction of social insurance for blue-collar workers by year.
Source: Global Work-Injury Database (GWIP) (Breznau and Lanver 2020)

Fig 1. visually demonstrates that after 1990 social insurance for work accidents covers most of the globe. Fig 1. visually understates the prominence of these laws. For example, in the U.S. where there is no national work-injury insurance, nearly all states have their own social insurance laws for work-injury.

The institutionalization of social risk in the hands of national policy appears widespread, but a careful look at coverage suggests it is perhaps not as strong or as widespread as welfare state scholars often claim.

In the Global North, the legal coverage of work-injury law is 69% of the labor force on average, whereas in the Global South it is only 37% (for those 169 countries for which data are available). There are some extreme outliers in the Global South with less than 6% coverage. As shown in Fig 2., many countries have very weak coverage as of 2010, despite having social insurance laws.

Fig 2. The legal coverage of work-injury insurance in 2010
Source: Global Programme Employment Injury Insurance and Protection Data (GEIP) (ILO 2014)

Given the prevalence of laws explicitly targeting the industrial, blue-collar workforce of so many countries, it is surprising that the coverage rates among the formal labor market are so low.

A main cause of this is lack of enforcement mechanisms. Many of these laws require extensive legal and governmental institutions. Infrastructure, bureaucracy, a system of enforcement thorough inspections, safe reporting possibilities for workers and measures that prevent corruption of funds or executions of justice tend to be weaker in the Global South.

Another reason is through clauses that all implementation on the whim of certain officials, additional laws that undermine national social insurance laws, or a lack of de facto punishments for not following laws.

Myanmar and Bangladesh are two cases that help illustrate these points. Both involve high density garment factory production. Myanmar enacted social insurance for workers in 2012, but the coverage for work accidents is less than 6% of the labor force. How is this possible when the law explicitly states “compulsory registration” in the social insurance system for employers? One possible answer is in the wording of certain clauses in the law. For example:

“The President may… exempt the regions which is not yet necessary to implement currently according to the plan to be implemented all or any part of the provisions contained in this Law or any establishment applied by this Law or any type of employer or worker.”

Myanmar Social Security Law, 2012, Art. (99)

Governments regularly grant exempt legal statuses to industries that are deeply embedded in global supply chains, industries that arguably would not be located in their countries if the cost of operation were too high. This might explain why Bangladesh did not bother to implement social insurance and instead relies on outdated forms of work-injury laws such as employer liability and a provident fund.

In 1980, the Bangladesh government passed the Export Processing Zones Act which created geographic areas that were no longer under the jurisdiction of national law and instead controlled directly by the president’s office. This meant that work-injury laws did not apply to all the ‘sweatshops’ feeding the global demand for inexpensive garments. With such an exception in place, the introduction of a social insurance law would not make any difference in coverage or addressing risk, unless it overrides the exemptions granted to special regions, those that can be granted in Myanmar on the whim of the president.

There have been horrific accidents in Bangladesh, with the collapse of the Rana Plaza as one of the worst in human history. This led to new coordinated plans between the industrial stakeholders, the Bangladesh government and the ILO, but it remains to be seen if the coverage rate increased from 12.5% – where it stood in 2014. Given the capacity for corollary laws to undermine social insurance laws, it remains doubtful. This suggests that the underlying normative notion of social risk and the government’s role as insurance arbiter it has not been institutionalized in Bangladeshi society and politics the way it has among many Global North countries.

With such low rates of coverage, the idea that social risks and welfare states are institutionalized at a global level may be overstated. Yes, laws are prevalent (see Fig. 1), but, no, they do not actually address the risks affecting nearly all workers (see Fig. 2) for a variety of reasons. As such they do not address social risk and may nurture sociological and path dependent institutions where work-related risks are normatively embedded as an individual or family risk rather than a social risk.

Data sources and replication files available on GitHub.

[1] Young coal miners (left), Bangladesh garment workers (upper-right), Uganda textile workers (lower-right)


Breznau, Nate, and Felix Lanver. 2020. “Global Work-Injury Policy Database (GWIP).” Harvard Dataverse.

ILO. 2014. “Global Programme Employment Injury Insurance and Protection | GEIP Data.”

Moses, Julia. 2018. The First Modern Risk: Workplace Accidents and the Origins of European Social States. Cambridge: Cambridge University Press.

Search OpenEdition Search

You will be redirected to OpenEdition Search