In 2022, Unions21 published the Using Data to Build Strong Unions report. A year later, I wanted to revisit this report to summarise its key points and discuss what, if anything, has changed.
It is important to note that this report is concerned with how unions use data, rather than the impact or risks of data collection from workers or everyday people.
The report also focuses on UK and European unions and case studies. However, the same institutional challenges (and opportunities) that Australian unions face are also present in the UK.
Why does data matter to unions?
I’ve argued for many years that unions were and should be data organisations:
the absolute, historic, advantage unions had to win was our monopoly of the “going rate”. This information and data advantage meant that we knew what the wage-rates of workers were, and we used that data to organise workers and build power.
This point is also made in the Unions21 report. Data can help unions build power for workers, increase member activism, and “match employers better”.
Data is relevant to the work of everyone in unions
The report emphasises that there was a “strong consensus that unions can should be using data more effectively in their work”.
A really clear example of this is in the report:
To give one example, a senior leader described how they had recently requested data on the rate of people returning to workplaces after lockdowns in the sectors they organise in.
They wanted this data not just to gain a better understanding of the current challenges facing members but to be more informed about the potential long-term implications for the union’s strategic aim to build density in those sectors.
What changes to organising practices will be needed if fewer people are in traditional workplaces? What changes will need to be negotiated to access agreements with employers? Will flexible working and fewer in-person interactions between colleagues change people’s expectations of their jobs and of unions? What happens to the concept of the picket line and the workplace rep when more people work remotely?
There is no part of a union’s work that wouldn’t benefit from more, better data, or that doesn’t already use data.
Even without digital databases, unions already collect and handle a significant amount of data — both internally and from external sources. Unions21 gives a range of useful examples:
- Employer data: company accounts, annual reports, information provided during bargaining, research from publishers like IBIS World
- Economic data: government statistics, recruitment data, jobs and inflation figures, consumer trends
- Membership data: from the membership system, surveys, email lists, pay agreements, case-work, workplace mapping, member/delegate training
Barriers to using data
The barriers highlighted in the report would sound familiar to any unionist in Australia:
- A lack of clarity about what data is, and
- A lack of familiarity with how it could be used
The second point could also include a lack of data literacy and digital literacy amongst union officials. This low data literacy remains a major barrier for unions to effectively use the data they have as well as develop vital new capabilities.
Another barrier is cultural and systems. These are interrelated. Reluctance by some inside unions to question ways of doing things and low willingness to change those processes is a barrier identified in the report. This is certainly the experience for many unions in Australia.
Overcoming the barriers
Unions21 makes the case for unions to train and recruit people with “advanced data competencies”, and to have a role that is the equivalent of “head of data” — a recognised senior manager with responsibilities that cover everything from finance, membership, internal union HR and even strategic research (such as extracting and analysing data from employers’ financial records), and with the authority and necessary resources to implement decisions across the whole union.
A role like this is clearly highly desirable, but possible only for larger unions.
Additionally, there is a case that expertise like this could be pooled across unions. For example, in Australia, unions who are partnered with the Union Innovation Hub could have a “shared service” of a member data analyst to assist with recruitment and retention.
Similarly, improving data literacy and digital literacy for union staff, officers and delegates is a major undertaking. Ideally, this would be led by union peak council training organisations like the ATUI. In the same way that we have a shared language and systems for organising, we need to embed a similar approach for data.
Data since 2022?
There’s at least two areas unexplored by the Unions21 report.
Data sharing between unions
The first is how unions can share data between each other. This is one of the major unrealised advantages that unions have — millions of workers are union members, and millions more are former members.
The vast majority of resigned members only resign because they change jobs (or retire) — not because they want to quit due to dissatisfaction. Unfortunately, unions make it hard to transfer membership.
Creating shared data standards is one way for unions to assist workers to transfer their union membership to a new union. (I wrote about data sharing between unions in 2019 here.) This capability would have obvious major strategic benefits for all unions.
AI and unions
The second is the import of machine learning/AI — an area of data use, collection and analysis that has massively advanced since the Unions21 report was written.
I’ve already written about how unions can use AI to grow — but it is worth expanding in reference to one of the use cases highlighted in the Unions21 report.
- Analysing the union’s central data repositories: Using a self-hosted gen-AI tool, unions could get summaries of existing unstructured knowledge from large libraries of PDFs and Word documents. This could also be used to summarise or pull information from online libraries of your union’s collective agreements, or years worth of member surveys.
(Note: Nicolas Blanc (from the union of managers in France) makes the point that generative AI systems can be installed on union-run servers or in private clouds. This means that many of the benefits of ChatGPT can be enjoyed without the problems of corporate surveillance and data-mining.)
Analysing and using the large quantity of unstructured (qualitative) data has always been a challenge for unions (and all organisations). The arrival of the large language models and generative AI goes a long way to providing a technical fix for that problem.
I recommend reading the full report from Unions21.
Understanding and using data effectively is vital for union growth and retention, building power for workers and advancing workers’ rights. Improving the data literacy and capabilities in your union will mean you can make better, more informed decisions for both strategy and tactics.
Presently, most unions are very far behind where we need to be in the “data arms race“.
Unions need to completely re-think the kind of organisations we are. Unions need to recapture the idea that we are information organisations, that we are data organisations.
Unions need to understand that our principle asset is data. Both data on members, but also data about workplaces, about wages, about industries, about safety incidents, about industrial compliance.
The Unions21 report is a good starting point for unions who want to do this.