Centre for Social Impact Swinburne

Transcript - Digital inequality in Australia: Results from the 2019 Australian Digital Inclusion Index

(CSI Swinburne's Social Impact Summer Webinar Series - Week 3)

Hi, and welcome to the Centre for Social Impact's, Social Impact Summer Webinar Series for 2019. I'm Rebecca Hutton, and I'll be facilitating today's webinar on the digital inequality in Australia.

Before I introduce our speaker, I'd like to make an acknowledgment of country. On behalf of those present, I acknowledge the Wurundjeri people of the Kulin Nation who are the traditional custodians of the land on which we now meet. I pay my respects to their elders, past, present, and emerging. I also pay my respects to all aboriginal and Torres Strait Islander people of Australia and hope that the path towards reconciliation continues to be shared and embraced.

OK. So today we are very lucky to be joined by Chris Wilson, who is the principal analyst for the Australian Digital Inclusion Index and research fellow here at the Swinburne Centre for Social Impact.

Welcome, Chris.

Yeah, thanks, Rebecca. Great to be here.

So today's webinar-- the way today's webinar will run is that Chris is going to start off by presenting an overview of the approaches to measuring digital inequality And then take us through some of the Australian Digital Inclusion Index results for 2019. This will take us through the first 15 to 20 minutes, and then we'll have 10 to 15 minutes for questions before we wrap up.

And you can submit your questions using the text function. And I'll be reading these out during the Q&A in the last part of the session. So now I'll leave you with Chris.

Great. Hi, everyone. Look, today you've got a lot to get through. What I'd like to do is provide an overview of digital inclusion, the Australian Digital Inclusion Index, and explore some of the key findings for the 2019 report. And I'm going to do that to illustrate some of the aspects of the level and nature of digital inequality that persists in Australia.

There's a few ways or few elements of this. The first one is going to be around defining and measuring digital inequality. We'll then look at the composition of the index and then we'll look at some of the results of the 2019 index.

OK, so like access to water or electricity, access to digital communication technologies are now essential to participating in almost all aspects of society. I mean, it's very hard to think of some form of social, cultural, or economic participation that doesn't incur some sort of-- or involve some sort of online activity.

Now while more and more people are engaging with digital technologies, there is a kind of issue around the network effects which means that for those that are online, there is rising value. But for those who are offline, there are rising costs. So to give you an example of this, I'm thinking of the banking sector. I mean, banking today is so much more efficient and effective than it was 15 years ago.

As a consumer, if you're online, you can do all sorts of sort of transactions, check statements, do all those things, transfer money, pay bills. They're amazing. They make our lives a lot more-- they bring huge amounts of value to our lives.

For those who are offline, I'd like you to perhaps spare a thought about the rising costs of being offline. What we've seen is the response from banks has been to start shutting branches and start having fee for service for over-the-counter transactions. And as a result of that, of course, if you're offline, the costs of being offline in terms of banking have gone up.

So there is this kind of, sure, more people are online. That's great. More people are getting value, but there's also a rising cost for those who are excluded. I really want to make no bones about the fact that social, economic, and cultural participation is undermined when we have digital inequality.

So digital-- kind of the concept of the kind of social impact of differential access to communication technologies is not new. This is something that's been going on for ages, and in fact, there's lots of work done around differential access to the telecommunication network. But really, it was in the kind of late 1990s when we saw the sort of diffusion of the worldwide web, mass diffusion across most developed countries that we started to really get concerned about the sort of notion of a digital divide.

And how this was conceptualized was really about people who had either access to the internet or no access to the internet. So the divide is really about that material access, a very simple kind of binary notion. And as a result of that, measuring was also quite straightforward.

So we have things like the falling through the net report, which is up in the top, corner here the cover there of anyway, which is a government report out of the US which starts to introduce this idea of a digital divide. It measures that by looking at household access to the telephone, to the personal computer. And to an internet connection. So again, it's an either or scenario. You either have access to you don't have access. You're one or the other side of the divide.

Now of course, over time, this attracted some criticism. Did having or not having access really explain the differential nature of the online experience? Was everyone who was online the same? Or were there differences amongst those who were online?

And so as a result, we started to think about not only what access meant. Did it mean just having a connection or did it mean the sort of type of connection you have or how often use that connection?

And also, those other elements such as, when you got online, what could you do there? What sort of skills did you have to make the most of being online?

And then, of course, the other really key aspect, I think, in this whole equation is how much does it cost? Is there a differential cost of participating in the online world?

Now of course, again, with this sort of move to a more complex conceptualization of digital inequality, we had to develop new measures. It wasn't, again, a single indicator of whether you had access or no access. How would you really determine just how included a person could be?

So the results of that kind of conceptualisation was thinking about producing a composite index which would kind of aggregate a whole range of variables around-- in the case of the Australian Digital Inclusion Index, as we'll see, access, affordability, and skills, and bringing those into a kind of a single number that would represent any individual's kind of aggregate level of digitally inclusion.

We're seeing this being developed in the international world, particularly through the International Telecommunications Union. I'll just put up a little picture there of measuring the information society. They've created these ICT development indexes. These have been going on for some time. That was 2009, I believe, that ICT Development Index.

In 2015, Telstra decided that he'd like to have an Australian version of this. And they were going into partnership with Swinburne University at first to develop the index, and that partnership now includes RMIT University as well as Roy Morgan Research.

So let's get onto what this index looks like. So the index is being built to include around 100 indicators of digital inclusion. These indicators are measured and then combined to form a numerical representation of an overall level of digital inclusion for an individual.

The indicators themselves, as I've said, come from effectively three key areas. Access, and here, we're not just talking about connection. We're also talking about the location of access, frequency of access, type, the amount of data you would have access to, so a range of indicators within that access sub-index, and then affordability, two measures there, relative expenditure, percentage of household income spent on access, and then the value of expenditure, how many gigabytes you're getting per dollar. Which again, if you think about it, if you're mobile only for instance, the amount of gigabytes you get per dollar is going to be significant less than if you have also, a fixed broadband of you think about the sort of data allowances that the fixed broadband offers.

The final sub-index and set of indicators comes from individual ability. These include not only online skills and activities you might be participating in, but also, your attitudes towards technology. Effectively, the way it works is you as an individual get a score on all of those 100 indicators.

Each of those indicators, the score is out of 100. We aggregate those together, you get an ADII score out of 100. The higher your score, the more digitally included you are.

So you've got you've got an index. Now you have to populate it with some data. We use the single source data set provided by Roy Morgan research. So it's about 15,000 respondents per year. And I just want to emphasize this index numbers on their own, that they really don't have any absolute value. It's not like it's something you'd count. It's about relativities. It's about comparing people over time, it's about comparing them in different locations, and comparing them in different cohorts.

And the data set we have for this is excellent, because the index starting now covers six years. So we can look back. Effectively, five years of changed since 2014.

The data is also available at national, state, and regional level so we can do some geographic comparison. And we have heaps of different population cohorts that we can look at with a whole range of demographic and socioeconomic factors so we can start to identify where inequalities exist. Let's have a look at some data.

All right, in 2019, the National ADII score was 61.9. Again, remembering as an absolute number, that's kind of meaningless. But where it has meaning is in that relativities about comparison. So here we look at the geography and we think at the state level, the ACT has a score of 67.6, much higher than the national average.

And we see in Tasmania, the score is 58.1. We have more and less included states and territories. And you can see here on the screen a range of other numbers for the other states in between those two and then those two outliers.

We can also track this over time. And we see here between 2014 and 2019, that the Australian Digital Inclusion Index national score has risen from 54 to 61.9. But a pretty slow start to 2016, and there's been sort of significant improvement since that time.

So we want to know more. Let's look at what's driving that. We see here the three sub-indices, access, affordability, and digital ability. What we really see here is a steady and large rise in access, particularly from, really, 2015. A big driver here has been the NBN, so the roll out of that is a fixed broadband infrastructure has driven some uptake of off the internet. It's driven high quality forms of access.

And also, of course, we've seen investment in the mobile infrastructure as well. So we're seeing some improvement in digital ability, a little bit a little bit patchy at times. We've seen some improvements in those areas. And there is quite a lot of investment in upskilling people in the digital area, so that's been heartening to see some improvement there.

Affordability is probably the area in which we have some concerns. We saw this drop away from 2014 to 2016 and then some improvements since then. The real key to the affordability issue is that, I mean, I think everyone will be aware that you're getting more gigabytes per dollar. This, I mean, particularly in the mobile space, has been a real-- really, really, the kind of value for money in that area has really gone up. The real issue is the amount of money we're spending on telecommunications and internet access.

So as a percentage of household income, we've seen a major rise there. For high income earners, probably not a big deal. But for those on low and fixed income, this is a significant issue and one I think really, what we should be looking at, looking at it at the federal level. And of course, seeing what we can do at state and local levels as well.

All right, so that's the national picture. Let's look at some of the socioeconomic contours of digital inequality. I'm now sending you a kind of quite a large graph here with a range of different indicators or cohorts that we look at in the report. Probably income quintiles is the most significant one there.

I'm going to look at that in some detail in a second, so I won't deal with that now. Employment is an interesting one. Obviously, if you're employed, on aggregate, your digital inclusion score is higher than the national average. And if you're not in the labor force, you drop away significantly. There's a lot of retired people in that particular cohort, but you've also got people outside the workforce in other capacities also in that group.

Unemployed is really interesting. We want to look at it too much detail today, but it's quite high, and that's because, of course, you've got younger people in that cohort, and they're looking for employment, so keeping those skills up. Obviously you need to have access to be involved in the job market. So getting close to the national average.

Education, it's a sort of drop away, which we might expect from those who are tertiary educated to those who have secondary education and those who didn't complete secondary school are in that less category. We'll look at age in some detail in a moment, a couple, a few other groups here at the end, those on disability, pensions, again, low fixed incomes, often suffering in terms of affordability.

And indigenous Australians, this is indigenous Australians in urban and regional areas. Significantly below the national average. And we have also, for a number of reports, if you'd like to look at those, done some work in remote communities. And we see quite interesting patterns of digital inclusion there as well.

So let's have a look at household income groups in some detail. Again, this is a major differential between those in the highest income quintile, household income quintile $150,000 plus against those in the lowest, at under $35,000. We're using household income groups because this is a good way to see the sort of resource access you having in a house.

 And if you use personal income, for instance, students who might live in a high income household but would have low incomes, would appear in a low income group, and that's, of course, not quite the sort of resource access they have. So household income is chosen for that reason.

So let's look at now what's generating this sort of low scores for the Q5 group. As we would expect, of course, it's affordability. It's that high amount of money going out of the household on internet access. And of course, the smaller amount of money coming in. Again, this is that this is a major concern and something that we need to think about, some sort of policy interventions around thinking about how we can enhance access for that group.

I mean, I think particularly in the age of digital transformation of government services, we know that these people on fixed incomes are often the people who are accessing those services. And of course, if they can't afford to do so, there's this sort of double barrier to entry to get to access income. And so that's, I think, a really key issue that we need to think about in terms of finding solutions. And of course, we see some lower levels of access and lower levels of digital ability as well.

Worryingly, we're seeing that the income gap between Q1 and Q5 over time over the five years since 2014 hasn't actually shrunk. So we're not really making improvements. There's been a digital inclusion improvement for both groups, but it's kept pace with each other. So in fact, it's not improving.

I'll have a look at now at age. And again, we see, perhaps as you might expect, for all the groups, we see a drop off in digital inclusion, particularly, of course, in the 65 plus group. Here, the issue is not so much affordability. I mean, of course, you can see here on the graph that affordability is an issue, but it is in that digital ability and skills area that we see a real drop away.

In that cohort of over 65, if you look at who's in that group, you see a really stock drop away as we get into the 70 plus, the 75 plus, the 80 plus. So it really does decline over time.

When we're looking here, I think, at the age gap, not the income gap. But here is the age gap. We've seen actually that the age gap has extended, so from 17.9 index points in 2014 up to 19.5. So again, not improving in this area, which again, is of some concern.

All right, I'm going to finish up now by looking at one of the case studies we did for 2019. This is very exciting. We went into Shepparton, and we did have a look at-- during a sample, a kind of convenient sample, so self-selected sample of recently arrived migrants who often-- from arriving on humanitarian visas. And we looked at this. We undertook a survey within the Digital Inclusion Index Survey, which allowed us to create scores for them and compare them to the averages.

What we see overall is that they track quite well against the Australian average. Now the overall kind of migrant community, the CALD community, actually does slightly better than the Australian average, and that's because there is a younger cohort in that group, so they tend to-- age tends to be a factor that improves some digital inclusion.

But let's have a quick look actually in that data, within that data set, and we see some really interesting things here, I think. So access is very high. Now if you've just arrived from overseas and you've got an-- and you need to communicate back home, obviously access over the internet, very, very important. So you're likely to have very, very good-- well, we've all spent money on getting good access.

Your abilities are usually fairly high as well. Again, younger cohort tends to have those higher levels of skills. But what we see here, of course, is again, you're arriving, you're on usually low, fixed incomes, and affordability is a major concern. So it's a nice little snapshot, some sort of specific research that we did this year.

We've done some other work in the remote indigenous communities that provide similar results, and we also did some work with the deaf and hard of hearing community as well. So some nice little spin-off pieces of research that are all in the previous editions of the Australian Digital Inclusion Index Report, which brings me to the report.

Let me just say that this is the 2019 version of the report here. And you can see here on the screen if you go to digitalinclusionindex.org.au, you will be able to find access to that report. And you'll also find access to the data that we use in your report and a little bit of extended data as well, so you can download that in PDF for Excel format so you can make your own slots and make your own graphs from that material.

So, look, that's an overview of the 2019 index. I think it's a good snapshot of where we are at in terms of digitally inequality in Australia. Thanks.

Great. Thanks for that, Chris. So just the start off the Q&A today, I have one question. So can you tell me, what has been the impact of the government's investment in the NBN on digital inequality?

OK, it's a really interesting question and one we get asked a lot because, of course, it was a fairly significant investment. So the index itself looks at different forms of fixed broadband access. And of course, you've got to remember that this was designed into sort of 2015, 2016, so it was still a fairly major form of connectivity. I think people were even using dial up at that point.

So we looked at the NBN and we thought, well, that's a higher quality connection. And I think if you look at the reviews of the NBN, of course we know there are pockets where it doesn't work, doesn't work well, there are individual issues. But on aggregate, it produces a more reliable and higher speed service. Now again, we're not suggesting that it couldn't be better, that there wasn't aspects of the proposal that were different and have changed over time, but on aggregate, it is better, and it does provide a better service.

So what we're seeing is with the uptake of the NBN, we've seen improvements in the access score one, because it's considered to be a kind of superior technology. And two, because, well, actually three, two other reasons. One has been that because after 18 months, people have to make a decision on that telecommunications access. When the window closes, when NBNs come in, they're closing off the old network.

There has been a transition of people who didn't have broadband before on to broadband, so we think that there's some sort of inspiring people to get online is one thing. And the other thing is taking legacy plans out of the marketplace.

So you might have been sitting there with your 200 megabyte plan that you've been on for 15 years, you've never changed it, and then suddenly you're forced to make that decision because of the close off. And then you're suddenly on to a multiple gigabyte plan. And that, of course gives you those data allowance benefits, so.

Yeah, right. Interesting. So we have somebody else with a question here. They're asking, what was the sample size for the Shepparton case study? It's interesting to see their score was higher than the Australian average. Can you tell me about that?

Yeah, sure. OK, so the sample size is about 170 for Shepparton. I mean, with all these things, you're trying to-- and again, it's a convenient sample. So we engage with local community representatives and ask for them to help us kind of bring together a community of newly arrived migrants.

So as I said, it was the overall sample. And we looked at the sort of population groupings there, and
we had some. We had a fairly big Afghan community represented, which we would expect in that locality and a few other communities involved in there as well. The second aspect to that question, I've just forgotten. But, so the second aspect of it-- oh, and it's interesting that it's higher than average. Yes. Yeah. It's high, I think. Well, I mean, yes, it does come as a surprise to some degree.

Except again, there's that real desire and need to be online. I mean, this is really is. I mean, with communicating back home, with being involved, obviously, in trying to access services in the job market. There is a real need to be connected and fairly well connected.

So there was a lot of-- there was there was very good forms of access. The real issue was around-- well, there's two issues, really. One was around affordability. And what we saw when we were doing the survey and actually participating in some of the actual field research and did some of these surveys with them, with local members of that community, what we found was that a lot of that getting really good value for money in their plans. They're really searching for the best-- the cheapest but best data allowances in the marketplace, doing really well. But again, it's that very low income combined with the necessarily kind of high fixed costs of getting access to the internet.

The other interesting part of that study was around language. We did do a separate little module on that. And what we really found there was a lot of people in that community don't have English literacy skills. And they actually, with all of them, I mean, most people think we just translate things into language, into local language.

But also we found that actually, lots of them don't have literacy in their first language. So this produces a really kind of interesting issue around online access. It's not just a digital literacy issue. It's around actually language matrices. And of course, we think there are some really great online solutions to that through text or voice translation, for instance, which we think offer some really good opportunities to sort of break down some of those other social divides that might be happening.

OK, thanks, Chris. We probably just have time for one more question. So can you tell me, is digital inclusion a generational problem? Will it disappear as those people who grow up with the internet become the older generation? That's an interesting question and one also that we do engage with quite a lot.

And it is difficult to tell. However, it looks to me from the data set that even people who have kind of been in the workforce when the internet has been around, and it has been around now-- we've kind of been internet active for 30 years. So it's not a new thing.

And of course, so we've had some time in sort of, well, 25 to 30 years. We've had some time to deal with this. And what we're seeing still at the older age groups who are kind of tipping into that 50 plus range, we're still seeing people kind of with declining or lower digital literacy. And I think that issue is around use it or you lose it.

So I mean, even I find that if there's some technology that I'm using for a while in my job and then I'm not using it anymore, then I lose those skills. The technology moves on, the software is updated, it's changed, there's new players in the market, and suddenly I'm confused about how to use that.

So I don't actually think that this is sort of generational issue because I think there is some aspects of that, but I don't think we should be holding out for some time and this problem will just naturally be solved. I think it's actually going to be a problem where we're going to have to continue to invest in. And digital literacy should be an ongoing form of investment by individuals, but also, of course, I think there needs to be policy interventions set that deal with that.

Yeah, OK. Right. So we're going to have to wrap up quite soon, but there's just another question that's come through.


If we can answer that one just quickly. So do you think the methods used to measure digital inequality are outdated?

OK. So I think just measuring access is. I think that just the notion of just home connectivity is an issue. Now although I wouldn't discount that entirely, I mean, I think the ABS Census has, in the last few iterations, included a question about whether anybody accesses the internet from home, from the dwelling, and I think that's useful, using any sort of device, whether it be over mobile, cell, broadband, or fixed broadband.

So I think that question, while it seems very simple, it at least gives us some idea of the sort of notion of those people who are not connected at the dwelling. So I think that's not the whole picture. And I think that's what the Digital Inclusion Index kind of says. It says that's really-- that's fine to know who's in or out, but those people are in the internet and on the internet, they're not all the same. They're not all having the same experiences.

Some people don't have the kind of skills and capacities to make the most of being online. Online access for some people means monitoring their data usage all the time, making decisions about not streaming things because they don't have the access. They don't have either the quality of access or they're worried about the cost of that. So I think that's a differential and leads to some people being included and some people being less included.

And of course, there's the affordability issue. So again, another aspect that's not captured by whether you have or don't have the internet, that kind of more complex idea of what kind of burden this is putting on you as a household. And again, we see, for instance, in various cohorts how that plays out.

Single parents, we did a little case study on last year. That was fascinating to see how people in rental accommodations who are moving more frequently than the population average don't tend to have fixed broadband. You've got to invest in that as a sort of notion I'm going to be here for some time.

So what do you do? You end up on mobile-only plans. What does that mean? Well, if you're a single parent, it means you're trying to provide access to multiple people in the household off of a mobile broadband device, be it a phone or some sort of mobile broadband or dongle. And the costs of that are significant, the cost per dollar are quite significant, so I hope that's answered the question.

I think it has. So I would like to say a big thanks to Chris for providing an excellent overview of the 2019 Australian Digital Inclusion Index.

Thanks very much.

So this is a third of our 2019 Summer Webinar Series. Please join us next week for our last webinar where Jo Barraket and Perry Campbell will discuss how social enterprises support health and wellbeing outcomes for young people. Thanks and good afternoon.

Thank you.