CSW Meeting 11: Innovative Data Approaches for Measuring Progress on Gender Equality and Women's Empowerment
Brief summary of presentation of information made
The interactive expert panel started with short presentations by five experts from national gender equality mechanisms, academia, civil society organizations and national statistical offices. This was followed by questions and contributions from Member States and other stakeholders addressed to the panellists.
Big data is increasing being explored as a way to close the gender data gap however there are problems with who owns the data and privacy. Big data cannot replace large statistical studies conducted.
Monitoring the implementation of the SDGs: The role of big data. Mr. Steve Macfeely, Head of Statistics, UNCTAD
Trying to monitor the implementation of the SDGs is enormous. For the SDGs there are 232 indicators. For the Millennium Development Goals there were only 60.
Current SDG indicators are broken into three tiers:
Tier 1: Concept, standards, data (We know what we are talking about and we have
Tier 2: Concepts, standards but no data (We know what we are talking about but we
don’t have data)
Tier 3: No concepts, standards or data (We don’t know what we are talking about and
we don’t have data)
Can big data help us with the SDGs?
What is big data?
There is no rigorous definition, depends who you are talking to. Big data is not defined by size. There are a number of features that make it unique.
- Variety – comes from a variety of sources – phones, cars, credit cards, searches, security surveillance, loyalty cards at stores, GPSs, online shopping – everything we do is converted to data
- Volume – is enormous from technology. Data is very unstable.
- Veracity – bias and noise – how is the quality? There is an inherent biases due to data divide – developing countries trying to provide data but it is much harder with fewer technologies.
- Velocity – speed- statistics are out of date before published, big data helps get around that
- Value – cost benefit
- Volability – quick technological change
Data volume is growing
- In 2007 - 300 exabytes data stored
- In 2017 - 2.7 zettabytes of digital data exist
- In 2018 - 16 zettabytes of data are produced globally every year
- By 2025 predicted to rise to 160 zettabytes annually
How much data are real?
- As few as 35% of twitter followers may be real people
- As much as 10% of activity of social networks may be generated by robotic accounts
- 11% of display ads and 25% of video ads are viewed by bots not people - 'fake clicks'
- 25% of reviews on Yelp are bogus
- 3% of Facebook accounts are fake and an additional 6% are clones or duplicates (equivalent of 270 million accounts)
- 2.5 quintillion bytes of data produced every day
Nearly every aspect of modern life is generating data but there are a lot of challenges
- The volumes are huge but a lot is junk – statisticians have to harvest through it all just to get something useful
- There has been a social change – one of the big changes is that we are generating the data ourselves – reviews, likes etc. - this is very different from historic data
- Growth in data is expected to increase exponentially. But we are recycling the same information, re tweets etc. How much is actually useful?
- We can’t take data at face value - a lot of Facebook accounts are fake. However there are huge opportunities.
- Digital data is global data, do we harvest individually or in one central place?
- Accreditation of data – the UN could help, can we accredit the work of NGOs and use it for official purposes?
- A lot of digital data not representative of the whole world.
- There are a lot of governance issues – ethical, legal, protection, privacy. How do we train statisticians and how do we disseminate the data.
- Ownership of data
- Reputational risks
- Data Divide
- Privacy (now and in the future)
- Data wars
- Oversight and Confidentiality
- IT and Cyber-security
- Quality Assurance
- Continuous Professional Development & Training
- Strategic Partnerships
- Communications and dissemination
Ubiquitous technology has created a deluge of digital data (and more to come in the future - Internet of Things, biometrics and behaviometrics)
Big data are still relatively new – so many norms and standards not yet set.
Big data simultaneously present both opportunities and challenges for official statistics.
Not clear that member states understand implications of big data – debate regarding SDG ‘data flows’ illustrates paradox.
Not enough thought has been given to ethics, privacy and governance.
Big data and the modernization of national statistical systems: Experiences of Colombia.
Mr. Jaime Sebastian Lobo Tovar, National Administrative Department of Statistics of Colombia (DANE)
In 2014 DANE began its process of modernization and innovation aimed at impacting the generation, production and dissemination of statistics in Colombia, The National Statics System allows us to improve the availability, quality, consistency and comparability of statistics and to reduce the duplication of efforts. We are working on the incorporation of non-traditional and Big Data sources into the statistical process: Smart Data, It is not only about how big but, how smart is its use.
We need to understand data from a legal perspective –private vs public, who is responsible for statistics?
What is the role of big data?
- Allows us to look at how alternative systems can enhance the systems already is place.
- It allows us to create systems of early warning and understand local, national and territorial levels.
- Big data can play a key role in gender analyses for example if we need to identify risk factors for women, how gender based violence occurs and allows public bodies to follow up and develop monitoring systems
Transforming the Data Revolution into a Revolution for Gender Equality.
Ms. Nandini Chami, Senior Research Associate, IT for Change
There are huge gender related data gaps in critical domains such as health, education, political participation and human security. Most gender indicators cannot be tracked due to issues of conceptual clarity, coverage, regular country production or international standards.
The logical next step seems to be capitalising on the ‘Big Data’ revolution around us, for “high-quality, timely and reliable data”. But it is important to tread carefully so that embracing Big Data does not mean submitting to the Big Data dogma
The 2 big data dogma
- The god view - that big data can capture everything and open up a gods eye view
- The end of theory - that we don’t need theory anymore
Big data limits
- Limit 1. The empty dream of complete representativity (crucial to understand in the context of the gender digital divide)
- Limit 2. The perils of ignoring contextual theory-building (important as gender norms operate in highly culture-specific ways)
How do we deploy big data?
- Pay attention to small data – it is important for fine tuning big data
- Guard against discriminatory/exclusionary results stemming from Big Data’s imperfections
- Ensure compliance with privacy and data ethics in Big Data partnerships
- Important to invest in data as a public good
The need for safe, open and gender responsive data.
Ms. Nnenna Nwakanma, Senior Policy Manager, World Wide Web Foundation
There is a real need for the need for safe and open gender data in Africa.
The birth of the internet brought about a global explosion of data. However, the divide between people who have internet and people who don’t is significant. The people most likely not to have access to the internet are women in poor areas. They are therefore often excluded from education, accesses to services and to be able to participate in democratic debates. It is estimated that 50% of people are offline – most of these are women.
Data has to adapt for people who are offline.
Africa is the first continent that heads of state want to understand the capacity of data
Data should be open by default – data that is legally and technically pen to everyone
There is an explosion of data mapping in Africa – We want to make data relevant to people
Data is no longer the holy grail of statistics
What is it that we need to do to make data an empowerment factor for women and girls? REACT
R - Rights – info and data is a right. We need to understand data and educate people how to.
A – Access – to the internet is important
C – Content – what do you find? What data is available? Is it what we need?
T – Targets – we need to measure it and evaluate ourselves
People need to have access to data that matters to them. For example women who collect shea nuts need to know how many women are dying from snake bites while collecting shea nuts.
The floor was opened to member states to ask questions to the panel. The following comments were made by the panellists during the question time.
- There is a huge demand for data and an emergence of new data suppliers who can produce statistics faster than official statistics. But what are the ethical principles?
- We need the UN to set up accreditation for NGOs who are collecting data
- Statistical literacy – we need to be literate in order to read it and make good choices on behalf of people.
- Need to integrate different sources
- Why are we collecting data – just to report? No we need to collect data to make a change
- Need to make data understandable and explain in a way people understand – especially rural people
- Data needs to be useful so we need to be making decisions of how to collect and what to collect and make sure we don’t have any gaps
- Disaggregated gender data – there is often not any gender component in data collection – how do we add a gender element to existing data. How do we integrate data to speak to each other?
- Administrative data – you need to liaise with departments and make sure they don’t change a data set without consultation. You have to help your government organise data in a way that it can talk to other data. If you talk to governments about stats they tend to fall asleep so you have to help them understand that they are an asset that help them make decisions. Data are infrastructure- it has to be taken seriously and be organised seriously.
- Open data – most data are proprietary. When we talk about open data we need to talk about opening all data – usually we just focus on governments but we have created all data so commercial data, data from social media, it all needs to be made open. At the moment statistic offices don’t have access to these.
What was of particular significance to share with The Salvation Army globally?
Measuring progress towards the Sustainable Development Goals is a difficult task which many NGOs, FBOs and UN agencies are struggling with. The Salvation Army has made a commitment to engage with and measure the progress it is making towards the SDGs. Understanding the use of big data could play an important role in this. It may be necessary to rely on data that is collected through technological platforms therefore understanding not only the benefits of big data, but also the limitations, is vital.