Measuring to Improve: A Company Perspective on Gender Data in Chilean Mining
For Carmen Duarte, gender-disaggregated data is not a box to tick but a tool for better decisions: as she puts it, what you don't measure, you cannot improve. With two decades in human resources and twelve years in mining, she has helped lead Anglo American Chile's work on diversity and inclusion through a period of remarkable change in the sector.
In this conversation, Carmen offers a candid company perspective on why gender data matters, and on how difficult it can be to gather reliably across different countries, cultures and legal systems. She traces Chile's journey from the lowest female representation among major mining nations to the highest, outpacing Australia and South Africa, and explains what it takes to turn data into operational change, from leadership pipelines to a programme that trained women in mining communities to drive specialist trucks. Her closing thought looks ahead to the harder questions still to come, on ageing workforces and neurodivergence, and to a sector she sees as restless to keep improving.
Measuring to Improve: A Company Perspective on Gender Data in Chilean Mining
Carmen Duarte, Anglo American Chile · 13 February 2026
This transcript has been lightly edited for clarity. The conversation is available to listen to above. Host: Laury Haytayan, MENA Director, Natural Resource Governance Institute.
When we talk about mining, oil and gas, we often focus on production numbers, revenues or national development. But behind all of that are people: workers, community members, suppliers. And no surprise, women and men do not experience extractive projects in the same way. Without gender-disaggregated data it is hard to fully understand those differences, and even harder to design policies or company practices that are truly inclusive. Frameworks like the Extractive Industries Transparency Initiative (EITI) recognise this, and the Standard now encourages countries and companies to pay more attention to gender. But collecting this kind of data, and using it meaningfully, can be complex, especially for companies operating across different regions, cultures and systems.
Carmen, please introduce yourself.
Thank you, Laury, thank you for having me. This is an exciting opportunity. I have been working in HR-related topics for the last 20 years, and in the mining industry for the past 12 years. I have focused mainly on organisational development, culture and leadership, but one of my strongest topics is diversity and inclusion. Chile has gone through a huge transformation in that area in the mining sector, so I have been very proud to have been part of that transition. I hope we have an interesting conversation.
When we say gender-disaggregated data, what do we mean exactly?
That is an interesting question, because when you sent me the invitation, I did exactly the same research: what actually is gender-disaggregated data? We tend to assume it is the same thing, or that we all understand the same thing. I think it is about having a gender perspective when you are looking at a trend, a decision, or a projection of something you want to achieve. You need data. And for things that point to more in-depth change, the gender perspective has to be present. For example, if you are talking about the communities you influence through an operation, you need to know the gender perspective: how many men, families and women you are influencing. So you could say it is data that is disaggregated by gender, but the point is why you use it: to make informed decisions.
From Anglo American's perspective, why is collecting and publishing this data essential, especially within the EITI framework?
There are two levels to consider. Anglo American is a multinational; it has representation across the whole world. Our main offices are in the UK, but we have a strong presence in South Africa and Brazil, in South America, and now a project in Finland. Being a global company, you are subject to scrutiny and to standards that have to be of the highest level. We have seen a trend in the industry: as well as being compliant in the information we release, in recent years there is a trend to have that gender perspective too, so we are subject to that. It also has to do with our policies, our beliefs, our purpose, which is to re-imagine mining and to impact people's lives. We believe that has to be done with diversity and inclusion. Then you have a second level, where Anglo American operates. I represent Chile, and in Chile we have a strong demand for gender-disaggregated data. So these two levels come together. That is why it is important for us.
Is the demand coming from government, or because Chile is now part of EITI? Where is the push coming from?
I think it has to do with a cultural change that has happened over the last few decades. Chile signing the EITI is a product of that. Chile has been very focused on closing the gender gap. It probably has to do with recent governments, with a state vision that we needed to incorporate this. There was a clear statement that we had a gap in gender representation, a gap in what women and men can experience in society, a gap in how we exercise our rights. So it was a government policy, but it also involves the public and the private sector. Speaking from the mining perspective, when I joined the industry in 2012 there were conversations about gender parity, and a realisation that we were one of the industries with the lowest representation of women. That started a conversation, and that conversation naturally had to open a new chapter on data, because you can have the purpose, you can want to close the gap, but you need data to take informed decisions. So it has been a journey in Chile, a public and private effort.
Do companies take practices from one country to another, or do they only comply with each country's rules?
If you are a multinational operating in different countries, you have to have high-level requirements, high-level policies and a high-level vision. Then you give countries the necessary autonomy to be compliant with their legal framework, while still following that vision, which for Anglo American is very clear: re-imagining mining and providing safe conditions. But when you come down to the specifics of a country, the alignment is to be compliant with that country's requirements. If there is a good practice in some countries, it is shared. For example, Anglo American is very good at tax management, so those practices are shared. You do a lot of benchmarking and good-practice sharing, and there are institutional organisations that help you do that. The ICMM, for instance, has standards that let you share good practices and work towards a standardised level.
What are the main challenges in gathering reliable gender-disaggregated data across operations?
It goes to two areas, which I would call technical and adaptive, or cultural. On the technical side, it has to do with your earlier question: what are the standards, what are we going to ask, what are we allowed to ask? There is very sensitive data when it comes to gender-disaggregated data. For example, are you allowed to ask whether a female worker is a carer for her elderly parents? In some countries you may not be. And why would you ask that? Because you can design good internal policies to support people who are carers for elderly parents. We know the population is ageing; we know Gen X carries a double responsibility of caring for both their elderly parents and their children. So it would be valuable, but you do not always know whether you can ask, and that depends on legislative or local requirements. On the cultural side, not everyone will see the value of gender-disaggregated data, the value and importance it can give a company. Someone could say, it does not matter whether the worker is male or female, we need standard reviews or benefits. But we know it adds a lot of value. So the challenge is at those two levels.
Can you give concrete examples of how gender-disaggregated data has informed decisions at Anglo American Chile?
Let me give an example from Anglo American Chile that weaves together the gender perspective and the community perspective. We have our workforce data; we know which are men and which are women, so we know whether we have the representation we want to promote. In Chile we are strongly focused on women in leadership positions, and we have a percentage there that is above the national average. We have developed a strong pipeline and tools to promote women into leadership. Now we are focusing on women who are operators and maintainers, the people actually on the mine site, where representation is not that big, and asking how we can promote it. We have data showing that, since we are in a main mining region, there are women available in the community to work with us, so we can design a programme to bring women into entry-level positions. Then you connect that with the communities. Anglo American has a strong tradition in Chile of connecting with communities, giving them tools to improve their lives and their environment, and technical tools to improve employability. We ran a programme targeting women in the Fifth Region, in central Chile, and gave them driving licences for special types of trucks to increase their employability. Now we cross that with our entry-level programme and invite those women to apply. So you connect different initiatives, based on data, both the data you hold internally as a company and the data you collect from the community, to make informed decisions.
How well does EITI's Expectation 9 support meaningful attention to gender, and should Chile or Anglo American push EITI further?
It is a new standard, so you have to test it to understand it. I could not yet say whether the Chilean industry will push beyond it, because we need to get to know it first. But what I can tell you is that Chile has been a very unique example in levelling up gender-oriented policies on female representation in mining. Let me give a little context. Chile is the main producer of copper in the world, and copper is very important for the energy transition, so we are a key player in mining. Ten years ago we were not a key player in female representation; we actually had the lowest standards of the big mining countries. Now, ten years on, we have the highest representation of the big mining countries, better than Australia, better than South Africa, who have done a lot of work in this area. That happened because we defined standards and policies, and because of public-private collaboration. There is sectoral representation, the Mining Council, which works with a think tank, Fundación Chile, that every year collects workforce data with a gender-disaggregated perspective. So we have that muscle. If you compare company sustainability or year-end reports from ten years ago to now, there is a lot more gender-disaggregated data; ten years ago I did not see much on the wage gap, for example. Multinational companies already have that perspective, so I think the EITI Standard will feel at home here and have a good reception.
What you don't measure, you cannot improve.
How do legal and technical terminology issues influence what companies are able to disclose?
It is key; the definitions matter. When you go to the definitions, the asterisks really matter. If you are talking about a wage gap, what is the asterisk? Is it total annual compensation? Is it before tax, or after tax? These are very small technicalities, but from the perspective of the person who has to provide the data, they are critical. And I do not think it is actually very standardised. I had the experience, before Anglo American, of working for a public mining company, which is under more scrutiny to be open with information because it is public. There was a request for information on the wage gap between the highest executives and entry-level positions. It makes sense to ask, but at that point we were not able to disclose it, because part of that information cannot be shared. So it would be easy to say, just give us that information, but the asterisk, the small line, is the key. It is a big challenge to have conventions, standards, and someone with the capacity to explain: this is the indicator we are chasing, this is the information we want. That will be key when Anglo American is asked for this type of information and has to produce these reports, across different countries with different legal requirements.
How can public-private partnerships help navigate constraints while advancing gender equity?
They are key, because the mining industry exists within an environment: the natural environment, the communities, and the relationships it has with the institutions that speak for mining companies. Through institutions like the ICMM we are able to present ourselves as an industry that is key for the energy transition and for the future of the next generations. These conversations help define standards, identify the topics that are key for our industry, and decide which indicators and standards we want to promote. The Copper Mark, for example, which is a very good standard for how mining is done, came out of conversations with these institutions. So it is a virtuous cycle. Conversations with institutions like yours help define the standards, and, more importantly, identify the standards we as an industry want to promote.
Looking ahead, what would help companies like Anglo American improve how they collect and use gender-disaggregated data?
It has to do with a well-known phrase in management: what you don't measure, you cannot improve. So you need to have the data on the table, and you need to go further in your "double clicks" of information. If this variable is present, what happens with another one? If we increase female representation in our workforce, what are we going to do about working shifts, about work-life balance policies? If you are going down this path of measuring, your initiatives have to be coherent with those measurements. And the challenges we face will go beyond gender, to other aspects of diversity and inclusion. We will have to do a lot more on ageing populations and on neurodivergence in the coming years, and those are very hard datasets to collect. So we have a big challenge ahead.