Does Better Internet Access Wind Up Disenfranchising Lower Income Groups?

Yves here. As counterintuitive as it seems, reducing the “digital divide” may not be the plus its advocates assumed it was, if the UK experience is any guide. But the mechanism that the researchers assume reflects prejudices about the poor. I suspect readers can come up with other hypotheses that would be consistent with the results.

By Alessandro Gavazza, Professor of Economics, London School of Economics, Mattia Nardotto, Research Fellow, DIW Berlin, and Tommaso Valletti. Professor of Economics at Imperial College Business School and the University of Rome “Tor Vergata” (Italy). Originally published at VoxEU

The internet is lauded for increasing access to information, but it is unclear whether this translates into a better-informed and more engaged voting populace. This column uses UK data to determine how the internet has changed voting patterns and aggregate policy choices. Internet penetration is found to be associated with a decrease in voter turnout, mainly among the lower socioeconomic demographic. Internet diffusion is also found to reduce local government expenditure, in particular on policies targeting less-educated voters. These findings point to a trade-off between the ‘digital divide’ and the ‘political divide’.

Successful democracy is underpinned by the ability of people to have their say; therefore, voter turnout is fundamental to sustaining the legitimacy of the democratic process. Voters’ information plays a key role in this process, helping to hold elected officials accountable to their electorate. The mass media is the primary source of information for voters, enabling them to monitor politicians and to use this information in their voting decisions. The internet has dramatically affected media markets, decreasing the costs of accessing information. At the same time, the internet has increased the availability of many forms of entertainment (such as movies, games, and social media), potentially inducing individuals to substitute away from news and from traditional media, thus crowding out their political engagement (Prior 2007). According to the Oxford Internet Survey (Dutton and Helsper 2009), over two-thirds of all British households have access to the internet, and 56% access it through broadband. Only 28% of those reported reading a newspaper online, however, and only 11% reported using the internet to look for information about an MP, local councillor, or politician.

Moreover, the extent of political participation affects aggregate policy choices. Large increases in suffrage provide, perhaps, the most interesting historical episodes – Lott (1999) examines the growth of US government spending as a result of women’s voting rights; and Lizzeri and Persico (2004) show that the extension of the voting franchise in 19th century Britain was associated with an increase in expenditures on local public goods.

The goal of our recent research (Gavazza et al 2015) is to empirically study the effects of internet diffusion on election outcomes, as well as on politicians’ policy choices. We exploit the dramatic growth of the internet in the UK through a rich dataset that reports the total number of local broadband subscribers in each node of BT’s local distribution network, matched with outcomes of local elections and with local governments’ policy choices (i.e. expenditures and taxation).  Access to the internet is often heterogeneous across individuals, and this variation is well-suited to studying the effects of the new media on electoral politics and government policy. Moreover, local government elections and policies provide an ideal laboratory to analyse these questions, since they usually exhibit greater variations than national ones, which depend on more-general issues (e.g. civil rights, security, or immigration). Finally, the circulation of local newspapers – the traditional source of information about local politics and local government – dramatically declined in the last decade as the internet spread, with potentially large effects on local elections.

Our analysis proceeds in two main steps. In the first step, we delve into the effect of internet penetration on electoral participation by merging our rich database on internet penetration to data on local election outcomes in each one of the 7,700 electoral wards in England over the period 2005-2010. Interestingly, evidence from the introduction of older, then-new media in the US shows remarkably different effects – Stromberg (2004) documents that radio increased political participation, whereas Gentzkow (2006) shows that the introduction of television decreased it. What about the internet?

Our empirical analysis faces one key empirical challenge – potential endogeneity concerns. Internet penetration is correlated with several observable demographic characteristics (such as income and education) that are also correlated with political participation. Hence, it is possible that some unobservable demographic characteristics could be correlated with both internet penetration and election outcomes.

To overcome this challenge, we exploit ‘random’ access to broadband internet across different geographic areas and over time due to weather shocks. Specifically, several regulatory and industry reports document that the weather affects the costs of providing reliable broadband.  Hence, we obtained monthly rainfall data from the UK Met Office, which we employ as the supply-side instrument that affects penetration across locations and over time through internet service providers’ costs. It is, arguably, uncorrelated with demand-side unobservables.

Figure 1 provides an example – it displays maps of two local authorities, one urban (Birmingham; top panels) and one rural (Sefton; bottom panels). In each panel, we also report the wards that belong to each local authority. The left panels display rainfall levels and the right panels display internet diffusion across electoral wards within the Local authorities. These maps provide simple graphical evidence of the variation in rainfall that is typical of the full sample; they also seem to suggest that a negative correlation between rainfall and internet diffusion persist across wards within these local authorities. While the maps only provide evidence of cross-section variations, we also exploit changes over time in the analysis.

Figure 1. Rainfall (left panels) and broadband diffusion (right panels)

valletti fig1 29 jan

Note: Rainfall (left panels) and broadband diffusion (right panels) across wards within the Local Authorities of Birmingham (top panels) and Sefton (bottom panels) in 2006. The black lines identify the boundaries of electoral wards.

Using this identification strategy, we find that a 10% increase in household internet penetration decreases voter turnout by 2.9%. This negative effect comes mostly from the lack of political participation among the lower socio-economic demographic – more-educated people still turn up to vote, whereas the political engagement of the less-advantaged people drops the most as they gain access to the internet. Moreover, we validate our identification strategy through several falsification tests that use data from local elections held before the diffusion of broadband, finding no effect (precisely estimated) on voter turnout in these pre-internet elections.

The findings of our first step complement the recent contributions of Falck et al (2014), who find that internet availability has had a negative effect on voter turnout in Germany; and of Campante et al (2013), who find that broadband had an initial negative effect on turnout in Italian national elections, but, over time, has fostered other forms of online and offline participation. However, the fundamental question in democracy is whether these changes in participation (due to the internet) have any effect on government policies. This is exactly the goal of the second step of our analysis, and the key contribution of our research.

To do so, we collect data on local authorities’ public finance choices (expenditures and taxation), and merge them with our internet penetration data, now aggregated at the level of each local authority. We find that a 10% increase in internet penetration decreases local government expenditures and property taxes by approximately 6% and 7.5%, respectively. Moreover, we find that expenditures that target less-educated voters (whose participation declines the most), such as expenditures on social housing and social services, decrease the most, whereas expenditures that target more-educated individuals (whose participation declines the least), such as expenditures on education, decrease the least.

Our findings corroborate the idea that internet availability has displaced more traditional media with a richer political content (i.e. newspapers). Moreover, they indicate that more-educated individuals tend to use the internet to get information and vote, while less-educated individuals tend to use if more for entertainment – they become less politically involved, and vote less. As a response to these differential effects, more-educated individuals become more influential in the democratic process, thus slanting policy choices more in their favour. The magnitudes of our results are sizable and, thus, our paper complements other recent contributions that find large effects of the media on politics (Martin and Yurukoglu 2014, Prat 2014).

In our view, these results give rise to at least two observations.

The first is that several countries have enacted policies to decrease the ‘digital divide’ by subsidising the supply of and/or demand for internet broadband, with the goal of decreasing economic and social inequality between different demographic groups.

  • Our results suggest that the use of broadband internet varies dramatically across demographic groups and increases the ‘political divide’ between groups.

The second observation is that many countries have recently increased the devolution of powers towards local governments.

  • Our results show that participation in local elections has dramatically declined in recent years, in part as the internet has displaced other media with greater local news content.

This raises questions about the accountability of these decentralised governments.

See original post for references

Print Friendly
Tweet about this on Twitter0Digg thisShare on Reddit0Share on StumbleUpon0Share on Facebook0Share on LinkedIn16Share on Google+5Buffer this pageEmail this to someone


  1. Foppe

    That’s all well and good, but given that the rise of broadband internet access occurred in the period in which the main parties decided to start pushing the same political/economic ideology, thereby making differences in election outcomes rather negligible, it is just as conceivable that we’re seeing that disfranchised folks are less inclined to bother to still vote (either because of better information or simply because they have more to do), while the disfranchised folks who (for whatever reason) don’t have internet access (and who may still read paper newspapers, and thus be more caught up in election theater) do. Either way, I’m not convinced that the explanatory framework provided holds water, if only because the authors have nothing to say about those larger socio/political developments.

    We’re seeing similar things here in NL, with many poor/older voters being derisive/near-hostile to the thought of voting (even for the socialist party, which has had no part in the developments of the past 20-30 years and which has criticized said developments from the get-go) because they’ve become so disillusioned.

  2. PlutoniumKun

    Really interesting results, I wasn’t expecting that at all. I wonder if there is an effect whereby the interactedness of the internet (posting comments, ‘likes’ etc) makes people feel they are participating in public life to such an extent that they don’t find going out to vote so important? Or perhaps it is simply a coincidental effect of internet penetration coinciding with the manner in which ‘third way’ politics has disenchanted many poorer voters.

  3. digi_owl

    Back when there was all this hoopla about the Arab Spring, and how it was Internet driven, i could not shake the impression that access to twitter and Facebook kept most people off the streets rather than on it. Why go out and talk to your neighbors etc when you could sit inside and watch things unfold from a distance? Shit didn’t start really hitting the fan until the bozos i charge started cutting the nation net connections, as then people had to step out onto the street and ask around.

    1. PlutoniumKun

      It was worse than that – the regimes quickly realised they could hijack social media to control the protests. Sometimes it was as simple as starting a twitter hashtag about meeting at a certain place to protest, and just arresting everyone who turned up. Plus it made it easier to track the ‘followers’ who were supporting but not turning up, and intimidating them – usually by arresting a family member. In reality, social media almost certainly made the Arab Spring less, not more likely to succeed.

  4. ambrit

    I agree somewhat with PlutoniumKun about the effects of Internet engagement. First, I am supposing that the “better educated” cohort of the population has ‘work’ to energize them psychically. Thus, this divide between self assured and goal oriented persons and the ‘precariat’ is masked by Internet use figures. Simply put, the poor have lots of time on their hands, and the Internet is, if nothing else, the best Time Sponge around. If the poor have nothing to live for, then that is exactly what they do live for, nothing. Also, an unemployed person has lower engagement with politics by being cut off from work related organizations. No Unions left to train them and lead them in socio-political actions. Finally, in a similar vein, the atomizing effect of the Internet effects the lesser focused and motivated people more than the focused groups. Given fewer external connections to provide motivating stimuli, the default position is ‘bread and circuses.’ The Internet becomes a subtle form of ‘house arrest’ for the poor and disenfranchised.

    1. tongorad

      The Internet becomes a subtle form of ‘house arrest’ for the poor and disenfranchised.

      I teach in a title 1 (economically disadvantaged) high school, and I believe your comment rings true. The degree to which today’s youth are screen-captured cannot be overstated. This is particularly sinister given all the tech billionaires that are driving so-called education reform.

      1. Gio Bruno

        I would never question the observation of someone on the front line, but most of today’s youth (rich and poor) seem infatuated with the digital small screen (cell phone and laptops). And who can blame them? Those small screens are ones they can control, as opposed to TV, and are quite useful in communicating with their friends.

        Do they overdo it? Probably, but so do I (and I’m not a Millennial). Reading really good books and having thoughtful direct discussion with others is very illuminating. As are the thoughts of the NC commentariat.

  5. visitor

    Not terrifically convincing, it seems to me that there are too many confounding factors that were not taken into account.

    a) The argument that Internet has displaced traditional media, hence reduced electoral participation, hinges on the rising availability of broadband Internet. How does the picture change when considering the period when Internet existed, but broadband was almost exclusively reserved for corporations and private persons had to be satisfied with narrowband Internet (dial-up, ISDN)?

    b) How much does broadband spread together with gentrification? Historically, high performance network connections (and the corresponding subscriptions) targets affluent customers first (ADSL when everybody else was on ISDN/dial-up, VDSL when everybody was switching to ADSL, optical fibre when others started moving to VDSL, etc). Since those affluent, heavy users of broadband are the ones who were less impacted by new policies, there is reason to investigate the usage of broadband across social categories and across time at a finer-grained granularity. It is not clear to me that the authors proceeded in this way.

    c) How much of that increasing usage of Internet was actually the result of public policies? I do not know about British boroughs, but there are countries where traditional forms of communication with the administration (i.e. paper or in person) have been replaced by interaction via Internet — e.g. people are forced to fill in their tax declaration on-line, or submit forms for a variety of procedures (and yes, it causes enormous problems because of unreliable Web sites, or for people who are too poor and old to deal with these technical hassles).

    d) If historical precedent shows that political participation decreased because of TV, and since broadband Internet has been substituting TV connections and programmes (Netflix & co), then perhaps this is not a new phenomenon but the continuation of a well-entrenched trend.

  6. JLCG

    I am in my eighties and have an easy life. I have been reading blogs for a long time and now I read fewer and fewer of them because the exposure of bloggers shows how superficial their views are, they repeat themselves, they preach but never report an act. It is very boring, so now I watch a lot of old movies on You Tube and cooking channels and listen to Bruckner and Beethoven and Wagner.
    The exposure to reality shows how shallow everything is. If I were farther from death than I am I would probably vote for those that would support my way of life.
    In a cynical way I might say that the Internet enlightens and the shallow tragic aspect of life is brought vividly into our minds.

  7. diptherio

    I’m baffled by what the amount of rainfall has to do with anything, and I fail to see the connection between rain and internet coverage that the maps supposedly display.

    Maybe when poor people get the internet, it just makes it all the more clear that their opinions really don’t count for anything. Maybe they did a web-search and found the British version of Gilens and Page and decided voting just wasn’t worth it. Maybe it’s just the fact that the neolibs/neocons have pretty much taken over and everyone with a lick of sense knows that the government is not there for their benefit – maybe people have just given up on the political process and the advent of the internet is totally unrelated.

    It seems that there is a serious lack of data here. I’m not at all convinced that they’re showing causation and not just random correlation. Pretty weak sauce, if you ask me.

  8. dk

    Agreeing with diptherio, weak sauce and the rain thing is ambiguous. And the idea that internet bandwidth should immediately (and solely) correlate to greater community involvement is faintly absurd; the negative correlation is equally suspect.

    On another level the philosophy behind the endeavor seems to be a sort of blame the messenger (or medium) idea. Why not reflect on the messages? Increasingly, internet traffic/usage is taken up by entertainment media streams (video and music). The current marketing thrusts of device vendors is that these devices are primarily for entertainment, not channels of general information. And broadband is often promoted in the same way. Additional uses/capabilities are rarely even mentioned.

    And connection bandwidth is not a sole or discrete factor. Let’s think about delivery platforms: hardware capability directly impacts use diversity, and design shapes usage as well. Platform display dimensions are shrinking, especially on the low end of the cost scale. Low income users will be more likely to 1) have only one device, and 2) that device will be hand-held (phone or small tablet). Interacting with volumes of text, either in large type with a lot of scrolling (and panning, depending on layout), or small type and a lot of squinting, is not more conducive on a small screen simply because the connection bandwidth has improved. Battery life may be a consideration for these users as well.

    In short, these devices are designed and promoted as entertainment delivery platforms. Increases in broadband access bandwidth does not osmotically produce a desire to engage in (non-entertainment) research.

  9. perpetualWAR

    Someone needs to do a little bit of research also about how internet applying for jobs disenfranchises the poor as well.

    I have a friend who used to make good money. She has been really devastated by the financial crisis and has not had a solid well-paying job for three years now. And to make matters worse, she lost her internet (due to cost) about a year ago. She told me the process she has to go through to apply for work. Because she lives in a rural area, she has to drive over 8 miles to the nearest McDonald’s, sit in the parking lot to access their Wi-Fi and hope the connection lasts throughout the application process.

    I was astounded. I never thought that the poor were even disenfranchised from applying for jobs in this digital age.

    Will life ever be fair?

    1. diptherio

      Every job that’s not minimum wage (and some that are) want you to apply on-line. We need community computer labs, like we have community gardens. Libraries fill the niche a little, but I’m guessing there’s demand enough for dedicated public computing facilities…

      1. ambrit

        Here in The American Deep South (TADS) our public library serves that function. The Downtown branch, (as far as that label can apply to an under 50K population nexus,) has roughly 50 computers in its’ public internet access plaza. I have seen those computers all occupied from time to time.
        The dreaded big box stores have learned this lesson. Most have a dedicated hiring portal machine.
        I’m always surprised, though I shouldn’t be, that computer literacy on the part of aspiring workers is assumed. “Better labour suppression through Science!”

  10. Synoia

    Successful democracy is underpinned by the ability of people to have their say; therefore, voter turnout is fundamental to sustaining the legitimacy of the democratic process.

    This is true:

    Successful democracy is underpinned by the ability of people to have their say

    This does not follow:

    therefore, voter turnout is fundamental to sustaining the legitimacy of the democratic process.

    It is an assumption. Unlike Switzerland, the UK is centrally governed, and its system is referred to as “Dictatorship by Parliament.” The voter have no say, whatever the Governing Party.

    A different hypothesis is that when people have more access to information, they discover their votes are meaningless, and abandon what they perceive as a useless process.

    A casein point is TPP in the US. It is being imposed by a central government, with a secret process, with no benefits to the populace at large.

    We are all familiar with the drop in manufacturing employment, well paying jobs, which has followed NAFTA and the inclusion of China in the WTO.

    For emphasis I will repeat:

    Successful democracy is underpinned by the ability of people to have their say; therefore, voter turnout is fundamental to sustaining the legitimacy of the democratic process.

    Only if the voters believe they have effect.

  11. Synoia

    Successful democracy is underpinned by the ability of people to have their say; therefore, voter turnout is fundamental to sustaining the legitimacy of the democratic process.

    This may be true:

    Successful democracy is underpinned by the ability of people to have their say

    This is an assumption:

    therefore, voter turnout is fundamental to sustaining the legitimacy of the democratic process

    As an alternative:

    Voter turnout falls because the voter s are better informed about how a central government ignores their wishes.

    An example: Single payer in the US.

  12. TimOfEngland

    I notice that the broadband and rainfall maps are dated 2006. In the UK between 2006 (10 years ago!) and 2012 there was a “broadband explosion” until then many people (rich and poor) were still on MODEMs! Now you rarely see a MODEM” back then they would have been better off comparing the number of subscribers to FreeServe (it was “free” if you were careful, profit was made via the cost of the phone calls). I sympathise a little with the rainfall link, my connections definitely became unreliable when it rained hard and I was paying about £150 month, being on the affluent side, for dual ISDN (128Kbps!). in justification I was in IT and was already running remote sys-admin services.
    However I agree with others that the corellation is weak.

    I believe the reduction in polical involvent is due to “Europe/EU” we are permantently informed (by theMSM) that the EU will pretty much ignore our prime minister and therefore our wishes and that we have to abide by the EU’s laws whatever.

  13. Synoia

    His assertion:
    If moar internet, then moar apathy.

    My hypothesis

    If moar internet and central indifferent government then moar disillusionment.

    Solution: decentralized government, like the swiss.

  14. McMike

    “This negative effect comes mostly from the lack of political participation among the lower socio-economic demographic – more-educated people still turn up to vote,”…

    government in the neoliberal age serves the interests of the higher income’d (in classical terms the ruling class) ..Why WOULDNT the ruling class come out to vote. They are “stakeholders” in the outcomes of electoral political contests in the neoliberal age..The working class will be stiffed no matter who wins and THIS has been pretty well empirically determined. The only solution for the working class (the lower income’d) is to lead its own mass , politically independent movement for a better deal, and I’m not talking about entering a ‘third party” electoral circus.

  15. Darthbobber

    “Disenfranchise” is pushing it. An endless stream of Kardashian and similar trivia no doubt contributed to a less-informed electorate, and so may the expansion of time-wasting ephemera on the internet. But there is no literal disenfranchisement. The internet’s growth and expansion has also been happening in an era in which poorer groups in the polity have increasingly felt that the political process just doesn’t respond, and I don’t think you can really isolaye a single factor (though they ATTEMPT to.)

    In any case, there’s a difference between contributing to a dropoff in people’s inclination to do something and preventing them from doing it if they are so inclined.

  16. Code Name D

    Interesting that no one brought in network and information theory. I have seen several studies that seem to identify access to the internet having the opposite effect of actually creating dis-informed public than an informed public. And what we know about network theory dose predict this behavior in advance.

    Issue number one has to do with access to information. Its counter intuitive but the internet actually crated additional barriers to correct information while giving grater access to incorrect information.

    Its rather bizarre but it appears that you-tube has seen a rapid boom in flat-earth videos. In fact there seems to be a revival of sorts to re-interpret the bible away from “globe earth” or “Nasa-earth” as they like to call it, to a more literal interpretation. It’s because disinformation is so readily available while selection algorithms tend to ignore any responses that might debunk misinformation. The problem is that in order to access information on the internet – you must first be aware it exists. And second is that there are absolutely no filters on the internet to intercept or regulate disinformation.

    The result is that the internet actually feeds into existing biases rather than offering meaningful information.

    The second issue is even more insidious than the first. I am trying to remember where I saw it, but there was a blog post where some one commented about the neo-liberal nature of face book. You like friends, adding them into your information network. You dislike others basically shadow-banning them from your network. The result is a group-bias generator that is inherently prone to authoritarianism.

    It’s not even really about banning the skeptics and silencing any and all decent. A skeptical perspective can only form an opinion based on the information presented to it. To have an informed opinion requires a cross pollination of information. But the nature of social media just doesn’t allow this. You are only preaching to the quire because only the quire has the ability to hear what you have to say.

    Eventually, conformity to the ideology takes place, even without skeptics knowing about it.

    Issue number three has to do with the raw and unrefined nature of the information you receive on the internet.

    For example, in the past, most voters learned about candidates through reporters that investigated candidates on their behalf and transmitted their conclusions vea TV or newspaper column. A professional has the ability to proves information down into a more refined form that the masses who are not reporters can make sense of.

    Today, such reporters are nearly extinct. Assuming the citizen even knows about the candidate to look them up on the website, they eventually have to replicate the work originally performed by the reporter, and likely heaving neither the time, experience, or training, are likely to draw false conclusion, assuming they even bother in the first place.

    The more complex the issue – the harder it is to sift through the data.

  17. washunate

    Does the phrase digital divide mean something different in the UK? In the US, that term is generally used as shorthand to describe the various ways in which lower income communities (particularly minority communities) are systematically excluded from aspects of society by having less access to technology generally and personal computing in particular in areas such as employment, housing, and education. The post doesn’t really touch upon that kind of inequality.

  18. Mr. Beer N. Hockey

    My working class friends and I swear we got ourselves a better grasp of our society without the aid of a computer in our first four or so decades of life than the vast majority of the best educated computer aided folks. We can all do real stuff too – without our fingertips. Like talk to our neighbours.

  19. Terry Flynn

    What about mobile (cell) coverage?
    3G coverage was rolled out quickly (in attempts to recoup huge outlays through the auctions) in UK – and the riots (Bristol etc) exposed that Blackberrys etc were used a lot in poor areas rather than fixed b/band.
    I fail to see how the weather is a good instrument in such circumstances.

Comments are closed.