Evaluating consumer perception and willingness to pay for broadband in Nigeria

Telecommunications Policy xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Telecommunications Policy journal homepage: www.elsevier.com/locate/telpol

Evaluating consumer perception and willingness to pay for broadband in Nigeria Nathaniel E. Uramaa,b,∗,1, Osita Ogbuc a b c

Department of Economics, University of Nigeria, Nsukka, Enugu State, Nigeria African Heritage Institution, Enugu, Nigeria Institute of Development Studies, University of Nigeria, Enugu Campus, Nigeria

A R T IC LE I N F O

ABS TRA CT

Keywords: Broadband Telecommunications Consumer Demand Utility Marginal willingness to pay

The study analysed the consumers’ perception, and willingness to pay for improvements in broadband qualities and attributes in Nigeria, using survey data collected in 2014 from 3300 respondent purposively sampled from eleven states in the country. The results show that 79%, of the respondents, still perceive broadband in Nigerian as expensive relative to its quality, which 58% and 84% of them conceive as slow and unreliable respectively. The respondents are, however, even willing to increase their payment for broadband if the qualities and attributes increase with an upper price limit of ₦10,000 per month. On average, consumers are currently willing to pay an extra 166% per month for an improvement from slow to fast broadband, 96% for an improvement from fast to high-speed broadband, 90% for the inclusion of Videophone and 36.2% for the addition of priority. These, in other words, indicate that their current perception of the price of broadband as expensive is not absolute, but relative to the existing broadband qualities, which inform the utility they derive from its usage. There is, therefore, need for improvements in the qualities and attributes of broadband in the country to boost its subscription.

1. Introduction 1.1. Background of the study Broadband network is increasingly being recognized globally as a very important proponent of economic growth and development. Parcu et al. (2011) report that broadband network is increasingly an integral part of the world economy, having a very strong impact on economic growth and development. Ernst and Young (2012), and Omigie, Zo, and Rho (2015) report that mobile broadband is one of the means through which developing countries can leapfrog into the post-industrial era. Okonji (2017) reports World Bank's finding that, “a 10 percent increase in broadband penetration yields an additional 1.38 percent increase in GDP growth for low to middle-income countries.” The direct effect of broadband on economic growth and development relates to the positive influence it has on business activities and transfer of knowledge. It also exerts a number of indirect effects through a variety of channels including empowering the workforce, speeding up adverts and easing the search for jobs (Jon, 2012). In line with these, OECD (2008) argues that broadband network will be as critical in the 21st century as the basic telephone line was in the 20th century. Due to this observed importance of broadband in economic growth, the Nigerian government has been making efforts over the years to promote its development and use in the country. Evidence of these efforts is the introduction of online services and e∗

Corresponding author. Department of Economics, University of Nigeria, Nsukka, Enugu State, Nigeria. E-mail addresses: [email protected] (N.E. Urama), [email protected] (O. Ogbu). 1 Director, Applied Economics unit, African Heritage Institution, 54 Nza street, Independence Layout, Enugu, Nigeria. https://doi.org/10.1016/j.telpol.2018.03.009 Received 1 July 2016; Received in revised form 16 March 2018; Accepted 19 March 2018 0308-5961/ © 2018 Elsevier Ltd. All rights reserved.

Please cite this article as: Urama, N.E., Telecommunications Policy (2018), https://doi.org/10.1016/j.telpol.2018.03.009

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

payment in almost all the sectors of the economy, approval of the 2013–2018 Nigerian National Broadband Plan (NNBP) that targets a five-fold increase in broadband penetration from 6% in 2013 to 30% by 2018 and auction of the 2.6 GHz spectrum frequency. In addition, the Nigerian Telecommunications Commissions (NCC) has put up some initiatives towards ensuring wider broadband coverage in the country. These include:

• the Wire Nigeria Project (WIN), with the aim of seeing that no place in Nigeria is more than 30 miles away from the backbone infrastructure; • the State Accelerated Broadband Initiative (SABI) project, which aims at encouraging the private sector to build and manage wireless broadband services, first, in the 36 state capitals, and then, other cities and towns in the country; • the Universal Service Provision Fund(USPF) that financed Community Communications Centre (CCC); and • the Rural Broadband Initiatives (RUBI); and the Rural Internet Service Providers (RISE) Complementing the above efforts by the government, all the mobile network operators in the country have been offering mobile broadband and data services as bundled products since the advent of smartphones, Blackberry phones, tablets, and wireless modems for laptops and iPad devices. Independently they took the initiative of deploying some terrestrial and undersea fibre links to the country. These include Globacom's US$700 million terrestrial fibre optic project to link Nigerian cities; MTN's US$300 million installations of a 3,500 km national fibre optic transmission network; and Airtel's construction of a 4,000 km fibre optic transmission backbone linking cities in the country. Included also is the partnership between Alheri Engineering, Phase 3 Telecom and Power Holding Company of Nigeria (PHCN) to roll-out 14,000 km of aerially deployed fibre optic cables over PHCN's Power lines (Jagun & Somolu, 2010). As a result of these efforts, a report from the Nigerian National Broadband Plan (2013–2018) shows that as of 2012 there are an appreciable number of submarine cable landings on the shores of the country providing over 9 Terabits/s of combined capacity. Uzor (2013) shows that, as at January 2013, there are four active undersea cable firms in Nigeria. These are the 7000 km MainOne submarine cable, valued at US$240 million; the 10000-km Glo-1 cable, with building cost of US$800,000; the NITEL's South Atlantic 3 (SAT 3) that is worth US$600 million; and the MTN's West African Cable System (WACS), worth US$600 million. In terms of a National Backbone fibre optic infrastructure, the report shows that most Long Distance Carriers (LDC) has among themselves, fibre presence in all the 36 states and the Federal Capital Territory (FCT) Abuja. It was also in their findings that while many routes in the country still do not have fibre coverage, there exists a proliferation of fibre along with some routes such as that between Lagos, Abuja, and Port-Harcourt. 1.2. Statement of problem Despite all the efforts described in the previous section, and the landing of appreciable number of submarine cables in the shores of Nigeria, there is still low subscription to broadband with high-speed capability for data and image downloads, video calls, video conferencing, video on demand, telemedicine, entertainment, home shopping or electronic commerce, Internet telephony and the likes in the country. Data from International Telecommunications Union (ITU) show that as at 2016, fixed broadband subscribers per 100 inhabitants in Nigeria is 0.01, while Ojobo (2017) shows that Nigeria's overall broadband subscriptions per 100 inhabitants in 2017 is 21. Compared with the 2016 World averages of 12.4 and 52.2, developing countries' averages of 8.7 and 43.6, and Africa's averages of 0.4 and 22.9 per 100 individuals for fixed and mobile broadband respectively (ITU statistics), it is clear that the subscription to both fixed and mobile broadband in Nigeria are still low. ITU statistics also show that as at 2016, fixed broadband subscribers per 100 individuals is 5.25 in South Africa, 5.20 in Egypt, 1.10 in Zimbabwe, 0.61 in Togo, 0.31 in Ghana, and 0.26 in Uganda. These figures show that even among African countries, Nigeria's fixed broadband subscription is still significantly low. Specifically, data on broadband demand shows that, though at same level as at 2001, Nigeria now lag behind, Kenya, Ghana, Zimbabwe, Namibia, Libya, South Africa, Morocco, Egypt, Algeria, Tunisia, Mauritius, and Seychelles in fixed-broadband penetration, making it the 204th out of 210 countries on ITU 2016 database. Some industry stakeholders and members of Board of Trustee of the Association of Licensed Telecoms Operators of Nigeria (ALTON) blamed Nigeria's low broadband penetration on weak demand for broadband capacities by Nigerians. Also, Uzor (2013) reports that, “despite the enormous bandwidth capacity emanating from the 7000 km MainOne submarine cable, the 10000 km Glo-1 cable, the NITEL's South Atlantic 3 (SAT 3), and the MTN's West African Cable System (WACS) cable systems, Nigerians are yet to benefit from the investments in terms of affordable and efficient broadband services. This report suggests that cost of broadband in the country relative to its quality may be contributory to its low demand and poor penetration in the country. Alongside the question of what determines the demand for broadband in the country, it is also necessary to address the question of consumers' willingness to pay for the broadband given the broadband qualities and attributes. 1.3. Objectives of the study Given the above facts, it becomes necessary to assess consumers’ perception of the current broadband cost and qualities, as well as how their willingness to pay is affected by the qualities and attributes of the broadband supplied in the country and their socioeconomic and demographic characteristics. In specific terms, therefore, the objectives of the study are to analyze: 2

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

1) Consumers' perception of the quality of the current Broadband in Nigeria; 2) Consumers' willingness to pay for broadband, given its current qualities and attributes; and 3) How their willingness to pay for broadband varies with improved broadband speed and reliability. 2. Literature review 2.1. Conceptualization and evolution broadband in Nigeria Recently, the Definition of broadband incorporates a number of qualitative indicators such as Class of Service (CoS) and Quality of Service (QoS). Hence, Jackson, Lookabaugh, Savage, Sicker, and Waldman (2002), and Atkinson, Correa, and Hedlund (2008) report that broadband is not about speed but rather, a medium that offers a wide potential set of applications and uses. They emphasized that the ‘broad’ in broadband should be recognized as meaning more than the ‘fat, fast pipe’, but should represent the nearly infinite possible uses and applications that might be developed, which a consumer might use. They concluded that broadband should be viewed holistically as a technical capability that can be matched to consumers' broad communication, entertainment, information, and commercial desires. According to Kelly and Rossotto (2012), … broadband is not just about improving the speed at which users can read online news, play video games, and engage in social networking … but an enabling platform that allows developers and individual users to enhance existing services and to develop previously unimaginable tools that improve business and society. In line with this, Business dictionary (2015) defined broadband as, digital communication technology in which the bandwidth of a single medium such as a wire, cable, or channel is divided into several independent pathways that enable fast and simultaneous transmission of different signals (data, audio, video), and interconnection of different devices on a network. Broadband can, therefore, be seen as a term used to describe high-speed, 'always-on' access to the Internet that enables one to view websites more quickly, enhances online shopping and banking, allows for the view, download, and sharing of large files such as photos, music, and videos. Broadband also enables access to new communication services delivered by voice over Internet protocol (VOIP), and to video conferencing. Although in 2015, the Federal Communications Commission raised the broadband minimum download speed from 4Mbps to 25Mbps, the Nigerian National Broadband Plan of 2013–2018 still define broadband within the Nigerian context as an Internet experience where the user can access the most demanding content in real time at a minimum speed of 1.5 Mbps. Currently, most of the broadband delivery mediums in the country are wireless technology. Okonji (2017) report that the widely publicized growth of broadband penetration in Nigeria is not a true reflection of the actual broadband penetration in the country since the growth is mainly driven by mobile broadband. It was introduced by the launching of CDMA mobile broadband networks using EVDO, followed by the UMTS in April 2007. The next was both from Glo and from MTN that launched HSPA services in February 2008, followed by Airtel's launching of its high-speed network in 2009. Jagun and Somolu (2010) show that as of 2009, 115,500 Nigerians had access to the Internet and of that number, only 67,800 people had broadband access. Okonji (2017) and Ojobo (2017) show that as of 2017, Nigeria was having a broadband penetration of approximately 21 percent. The paper, however, acknowledged, “with the International Telecommunications Union putting fixed broadband penetration in Nigeria at 0.01 percent, the bulk of this broadband access has been through mobile broadband, which does not reflect the true broadband penetration level of a country.” According to MainOne Cable Nigeria LTD (2013), the high growth witnessed in the ICT sector in Nigeria has been largely restricted to the voice telephony segment and failed to extend to broadband and data segments, amounting to a huge dichotomy between the voice telephony and data segments in the country. 2.2. Factor affecting the demand for broadband Pricing is a very important aspect of demand analysis in general and IT services demand in particular. Hassett and Shapiro (2009) and Jameel, Ehsan, Mirza, and Bhatti (2012) for example, show that price is the strongest determinant of a broadband subscription. Supporting this, Rappoport, Taylor, and Kridel (2003) show that at $20 per month, a 10 percent increase in the price of broadband reduces demand by 5.3 percent, a price elasticity of −0.53, while at a price of $50 per month, a 10 percent price increase reduces demand by 9.8 percent. Duffy-Deno (2000) reports that a 10 percent increase in the price of high-speed connections in the year 2000 reduced the demand by 10.8 percent overall, but by 15.9 percent among those with income below $25,000, and from 8.5 percent to 10 percent for all other income groups. Jackson et al. (2002) report availability, price, application, content, and externality as the major driver of demand for broadband. Goolsbee (2006) after controlling for individual demographic characteristics show that price elasticity of demand for cable Internet range from −2.8 to −3.5%. Rappoport, Kridel, Taylor, Duffy-Deno, and Alleman (2002) using logit model and a sample of 5255 households from the USA in 2000 show that broadband own price elasticity for Digital Subscriber Line (DSL) is −1.46. Varian (2002) estimates willingness to pay for bandwidth speed using a reduced-form demand for bandwidth and reports own-price elasticity ranging from −1.3 to −3.1%. Some studies also show that location of the consumers may affect their demand for broadband. Along this line, Fairlie (2004) found a negative correlation between rural location and the likelihood of subscribing to Internet services in the United State, while Savage and Waldman (2009) show that both the urban and rural households value improvement in bandwidth size and are willing to pay about $8 and $25 per month for low and high bandwidth respectively. Wareham, Levy, and Shi (2004) in their study to find out 3

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

the socio-economic factors that affect the diffusion of the internet and 2G mobiles in the USA show that Mobile adoption is positively correlated with income, occupation and the individual's residential area. Manzoor (2014, pp. 1–17) highlighted service quality among the factors affecting the demand for broadband in India. In Kenya, Mugeni, Wanyenbi, and Wafula (2012) show that improvement in digital literacy and security of broadband are among the significant factors affecting the broadband demand. In Nigeria, Umezurike et al. (2015) report that Poor Power Infrastructure, Pricing, and Right of way are the main factors affecting the broadband penetration in Nigeria. 2.3. Price of broadband services A report by ITU (2012) shows that from 2008 to 2012, fixed-broadband price as a share of gross national Income (GNI) per capita dropped by 82%, representing 1.7% of monthly GNI per capita in developed countries, and 30.1% of average monthly incomes in developing countries. This according to the report was still high in developing countries. Comparing the cost of fixed and mobile broadband in developing countries, the report shows that mobile broadband services cost considerably less than fixed-broadband services, with 18.8% of monthly GNI per capita for a 1 GB post-paid computer-based mobile-broadband plan compared to 30.1% of monthly GNI per capita for a post-paid fixed-broadband plan with 1 GB of data volume. Existing studies also reveal that reliability of the broadband network is a major factor affecting its demand, and the consumers’ willingness to pay. Though not the only determinant, they show that price and demand for broadband are highly related to its reliability and data transmission speed. Rosston, Savage, and Waldman (2010) for example, show that a representative household is willing to pay about $20 per month for more reliable service and $45–48 for an increase in speed. The reports show that in the USA, a representative household would be willing to pay about $59 per month for a less reliable Internet service with fast speed (Basic); about $85 for a reliable Internet service with fast speed and the priority feature (Premium); and about $98 for a reliable Internet service with fast speed and all other activities (Premium Plus). Although high speed is one the qualities that distinguish broadband from the Internet, FCC (2012) reports a disparity in the advertised and actual speed of broadband services. The advertised speeds are the ‘Up to’ download and upload speeds stated by the providers during their advert and in their price chart while the actual speed is the actual rate of average download and upload speed experienced by customers. The report shows that, since different broadband providers may not use the same methodology for determining their advertised speeds, they are likely to vary on how well-advertised speeds match actual delivered speeds. Confirming this, Ofcom (2012), from a UK broadband study in 2011, revealed an average advertised speed of 16.3 Mbps with an average actual speed of 7.6 Mbps, a gap as large as 8.7Mbps. This made the Advertising Standards Authority (ASA), and the Committee of Advertising Practice (CAP) to issue guidance that took effect from April 2012 that broadband providers should advertise a given broadband speed only if at least 10% of the customer base can achieve it. Contrary to the UK scenario, a report on consumer wireline broadband in the USA shows that Internet Service Providers (ISPs) deliver on average, 96% of their advertised speed during peak intervals, with five ISPs, routinely meeting or exceeding their advertised rates (FCC, 2012). 3. Methodology 3.1. Theoretical framework The characteristics of Broadband technology changes with time hence, evaluation of its demand involves evaluating how its demand responds to the rapid changes in its attributes and services. Addressing this, Savage and Waldman (2009), and Rosston et al. (2010) outlined a simple theory of optimal choice for the broadband attributes and services, which they used to explain why highability consumers value bandwidth differently. Consumers are assumed to maximize utility function of consumption(C ) , leisure(L) , broadband bandwidth(b) , and time online(t ) , subject to monetary and time constraints.

Maxh, b, t U (c, L, b, t ) s. t . c = y + wh − pb b − pt t L = T − h − t − T (h, b, t , a)

(1)

where U is utility, y is non-wage income, w is the wage rate, pb is the per unit price of broadband, pt is the per unit price of time online, T is total time available, and T is the essential time represented by the production function T (h, b, t , a) , with h , the number of hours worked and a, an efficiency parameter that reflects the technical ability of the household. T is convex in b and t , while b and t are complements in production such that increasing b will raise the marginal productivity of t . Similarly, a augments the productivity of b and t , decreasing the time for a given input level. Utility is concave in c, l, b and t , so that Uc , UL, Ub, Ut < 0 and Ucc , ULL, Ubb, Utt < 0 , and the desirability of leisure increases with consumption so UcL > 0 . The above theoretical model shows that individual consumers may have heterogeneous preferences toward bandwidth. For the ith respondent facing the jth choice experiment, the conditional utility for broadband service alternative k is modelled as:

Uijk = βij′ Zijk + (βs + αi′ δ ) bijk + βp pijk + εijk

(2)

where Zij is a vector of quality attributes and services, αi is an m×1 vector of indicators for an individual's technical ability, bij is the bandwidth, pij is the price of broadband services, and the elements of β and δ are preference parameters. Defining x as: 4

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

⎛ Zij ⎞ ⎛ βZ ⎞ ⎜ bij ⎟ ⎜β ⎟ x ij = ⎜ β=⎜ S ⎟ αi bij ⎟ δ ⎜ p ⎟ ⎜β ⎟ ij ⎝ p⎠ ⎝ ⎠

(3)

and inserting equation (3) into (2) will give:

Uijk = βij′ Zijk + εijk

(4)

3.2. Model specification In order to analyze the consumers' willingness to pay for broadband, this study adopted the Savage and Waldman (2009) model of equation (4). However, it used the consumers’ marginal willingness to pay (WTP) as a measure of their utility and a dummy variable to capture the differences in status quo broadband attributes and services (x) from the hypothetical one. Hence, the model of equation (4) will be: n

WTPij =

k

∑ ∑ βj Xij + εi

(5)

i=1 j=1

As stated in equation (3), the variable x is a vector of broadband attributes and services that includes speed, reliability, priority, and access to home wireless, movie rentals, tele-health, and videophone. Speed and Reliability are standard features of all current Internet services. The rest are the activities or the demanding contents, which together with a speed of over 1.5 Mb/s forms a broadband. With respect to speed, Internet connection can be ‘slow’, ‘fast’, or ‘very fast’. A reliable broadband connection according to international standard is a broadband connection with less than 1-h disruption per year. In other words, it means a broadband connection with up to 99% efficiency. Priority is a broadband attribute that enables the subscriber to designate some downloads as high-priority, so that they travel through the Internet at relatively faster speed. Home wireless is a broadband service that enables the subscriber to connect a device like laptop, tablets, phones, etc. to the Internet wirelessly within a given distance from home. Movie Rental is a broadband service that enables the subscriber to download high-Definition movies and TV shows, while Videophone is a broadband service that enables the subscriber to place free video phone calls over the Internet. The above model assumes that the change in marginal willingness to pay for an improvement/additional broadband attribute or service is constant across the consumers’ socioeconomic and demographic characteristics. The reality, however, is that the broadband consumers do not have identical preferences for a given improvement in the broadband attributes and services. In line with the findings of Hassett and Shapiro (2009), household demand for broadband and hence, their marginal WTP for broadband is influenced by their socioeconomic, demographic and other characteristics. For instance, an educated consumer may have a high preference for an increase in broadband speed than consumers with less or no formal education. In order to find out the impact of socioeconomic and demographic characteristics of consumers in their marginal WTP for an improvement in broadband quality, the assumption of homogeneous consumers is relaxed by including the socioeconomic and demographic characteristics of the consumers in the model of equation (5) to obtain the model of equation (6) below. n

WTPij =

k

∑ ∑ βj Xij + φi Di Xij + εi

(6)

i=1 j=1

The variable Di is a vector of dummy variables that takes the value 1 if the respondent has certain socioeconomic and demographic characteristics, and zero otherwise. These sets of characteristics include having awareness of broadband, living in urban area, having formal education up to a certain level, being a public servant and living below the poverty line. From equation (6), the marginal WTP for a consumer that possess a certain socio-economic characteristics for an additional broadband attributes or quality is given as pj + φ, where that of those without such characteristics is given by pj , making φ the difference. The selected consumers' socioeconomic and demographic characteristics are broadband awareness (Aw); having up to secondary education (Edu); place of residence (urban/rural) (Urb), and income level (Poor). Being categorical variables, Aw takes the value 1 if the respondent has a deep knowledge of broadband and zero (0) otherwise. Edu takes the value 1 if the respondent has up to secondary education and zero (0) otherwise. Urb takes the value 1 if the respondent lives in urban area and zero (0) otherwise. The pub takes the value 1 if the respondent is a public servant, zero (0) otherwise, while Poor takes the value 1 if the respondent's income is below the poverty line, zero (0) otherwise. Objectives 1 and 3 namely, analysis of consumers’ perception of the current broadband quality in Nigeria, and how their willingness to pay for broadband varies with improved broadband speed and reliability were addressed using descriptive statistics. 3.3. Data source and survey design The analysis used primary data from a survey conducted for the Nigerian Communications Commissions in 2014. The survey was conducted in ten states in Nigeria and the Abuja namely: Lagos; Ogun; Rivers; Akwa-Ibom; Anambra; Enugu; Plateau; Kaduna; Kano; Bauchi; and the Federal Capital Territory, Abuja, hereafter, assumed to be a state. The eleven states were made up two states from five 5

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

of the six geopolitical zones in the country namely: South-East; South-West; South-South; North-Central; North-West, and one from North-East due to high-security risk in the North-East region as at the time of the survey. These states were selected based on the availability of broadband Internet and the population density, which determine the current broadband deployment in the country. Since the main aim of the study was an evaluation of consumers’ perception and willingness to pay for broadband, it was important that most of the respondents were from areas where there are broadband connections (not necessarily using broadband). To ensure that this was the case, the states were further stratified, and areas with broadband access selected, including the tertiary institutions in each state. All the states had a sample size of 280 each except for Lagos that had 500 due to its high population size and wide broadband coverage. In each state, the sample size was divided 40:50 between the tertiary institutions in the state and other sampled area that were mainly urban areas. Finally, the interviewers used random sampling technique in each of the selected areas and tertiary institutions to locate the respondents. A maximum of two respondents between the ages of 18 and 65 was interviewed in each of the selected households. The survey was a face-to-face interview where the interviewer asked the respondents questions from a well-designed and tested semi-structured questionnaire and recorded their responses. The survey lasted for two months, taking place simultaneously in all the states. The average response rate was 97 percent, but the strategy used by NCC contractor to ensure that they got up to the recommended sample size was to administer more than the required size. For instance, the survey covered 300 respondents in each of the other states and 550 in Lagos. The excess took care of nonresponses and questionnaires with incomplete and inconsistent responses, which were filtered during validation. The crosschecked and validated questionnaires were coded in SPSS version 20 and the coded data transferred to Stata 13 where all the analyses were made.

4. Research results and interpretations 4.1. Description of the survey population As contained in section 3 above, the survey was conducted in 11 states in the country, and all the states had a sample size of 280 each, except Lagos that had 500 due to its high population size and wide broadband coverage. These together gave a sample size of 3300. After editing the survey data, however, the total sample used for this analysis came down to 3253. Since the survey is to evaluate the demand for broadband in Nigeria, and broadband is still concentrated mostly in the urban areas, the survey was designed such that 82.79% of the respondents are urban residents, while the remaining 17.21% are semi-urban and rural dwellers. The age distribution of the respondents is as shown in Fig. 1 below. The chart shows that 57% of the sample is within the age of 25–35 years, and those within the age of 18–24 years followed it with 27% of the sample. Those within the age of 36–45 years constitute 14% of the sample, while those that are above 45 years constitute only 2% of the respondents. The descriptive statistics of the sample further show that 36.2% and 34.4% of the respondents are paid employees and students respectively. The self-employed constitute 23.2% of the sample, while the remaining 6% are unemployed. Forty-nine percent of the employed respondents are in the public sector, and 51% are in the private sector. Education wise, 82% of the respondents attended or are in a tertiary institution, 17% have a secondary education while the remaining 1% attended only primary schools or have no formal education. Sex distribution of the respondents is 66% male and 34% female. The average income of the respondent households is NGN90640.19, but with a minimum of NGN3000 and a maximum of NGN7000000 and a high standard deviation of 343784.8, showing the high inequality of income distribution in the country. Although the survey was restricted to areas with access to broadband, some of the respondent households reported that they were not using broadband. The descriptive analyses, however, show that up to 2991 households, which is 91% of the sample use broadband as at the time of the survey.

Fig. 1. Age distribution of the respondents. Source: Authors' computation from the survey data 6

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

Fig. 2. Consumers' perception of current broadband cost and speed in Nigeria.

4.2. Assessment of consumer's use and perception of broadband cost, and quality The consumers’ perceptions of the current cost of broadband were rated in 6-point scale, with very cheap and very expensive at the lower and upper extremes respectively. The results in Fig. 2 (left) show that only 2% of the sample (3253) viewed broadband price as being very cheap as at the time of the survey. Thirteen percent perceive it as cheap, while 16% see it as moderately cheap. These summed up to 31% of the respondents that perceive the current broadband as being cheap, against 69% with the perception that the current broadband is expensive. Out of this 69%, 31% view it as moderately expensive, 27% as expensive while the remaining 11% view it as very expensive. Hence, the majority of the broadband subscribers in Nigeria still consider it as being expensive. Information from the survey showed that the average monthly expenditure on broadband per respondent is ₦6500, with a minimum of ₦1000 and a maximum ₦80000, but with large variations. In line with this, Paradigm Initiative Nigeria (PIN) in its Policy Brief number 4 reports that “Broadband remains unaffordable to the vast majority of Nigerians as fixed broadband plan costs 50% of national minimum wage while mobile broadband plan cost 17% of the national minimum wage”. The consumers’ perceptions of the speed of broadband were equally ranked on a 6-point scale, from very slow to very fast. The results of the descriptive analyses as contained in Fig. 2 (right) above shows that 41.3% of the consumers reported their current broadband as fast. Within this group, only 8% perceive their current broadband as very fast, 7.2% believe it is fast, while 26.2% of the consumers perceive their current broadband as being moderately fast. On the other hand, 58.7% of the consumers perceive their current broadband as being slow (30.5% moderately slow, 17.7% slow, and 10.5% as very slow). At the state level, the percentages of the broadband consumers who believe that their broadband is fast are shown in Fig. 3 below. Lagos state has the highest number, 62.7%, followed by Abuja with 55.1%, and Enugu with 53.4%, while the least two states are Plateau and Bauchi with only 19.6% and 18.5% of the respondents respectively. Moreover, Fig. 4 below contains the descriptive analyses of consumers’ perception of the reliability of their current broadband network. Reliability is defined in this context as broadband service that is rarely disrupted by service outages. The reliable broadband Internet may go down only for less than an hour once or twice per year. In other to analyze the reliability of the current broadband services in the country, the consumers of broadband were asked how frequent their broadband goes down within a specific period, and the result is as contained in Fig. 4) below. The pie chart of Fig. 4 reveals that only 16% of the consumers of broadband in Nigeria as at the time of the survey saw it as being reliable. Out of this 16%, only 1% of the sample perceives their current broadband as very reliable, 5% perceive it as reliable while 10% perceive it as moderately reliable. The remaining 84% of the consumers perceived the broadband in the country during the survey year as unreliable. Out of these, 24% see it as very unreliable, and up to 37% report it as unreliable. In general, therefore, most of the broadband consumers in the country have the perception that the current broadband in the country is still expensive, slow, and unreliable. The study further analysed the type of activities for which the consumers use their broadband to give a better understanding of

Fig. 3. Percentage of the respondents who reported their broadband as being fast. 7

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

Fig. 4. Consumers' perception of the current broadband reliability in Nigeria. Source: Authors from the Survey data

their perception of broadband cost and quality in Nigeria. The results show that 44% of the respondents use their broadband mainly for academic/education activities, 35% use theirs mostly for leisure and entertainment, while 21% of the respondents use their broadband mainly for profit-oriented business activities. 4.3. Marginal willingness to pay for broadband in the country In order to address the second objective of this study, the consumers' willingness to pay for broadband in relation to broadband quality and attributes were analysed. The result in the Table 1 below shows consumers' willingness to pay per month for an upgrade from all you can eat package without given attributes or services to that with such attributes or services. In order to avoid the biases that will result from the assumption of a linear relationship between willingness to pay and speed (slow to fast and fast to very fast), both fast and very fast with slow as the base were introduced in the model. The result as contained in the Table 1 assumes that the consumers are homogenous in their socio-economic and demographic characteristics and shows that all the attributes and services have a positive impact on consumers’ WTP. The extra amount in ₦/month the consumers are willing to pay as a percentage of the original amount, for an ‘all you can eat’ package with the inclusion of any of these attributes and services, except for priority and telehealth are statistically significantly greater than zero at 5% level of significance. For example, on average, consumers are willing to pay an extra 166.3% per month for a broadband package improvement from slow to fast with other attributes and services remaining unchanged, and an extra 96.4% percent for a movement from fast broadband network to very fast broadband network. Taking the consumers WTP as a measure of their preferences for broadband attributes and services, the consumers’ preferences for broadband attributes and services in decreasing order of importance is: fast, very fast, reliability, movie rental, videophone, home wireless, telehealth, and priority. On average, they are willing to pay extra 166% in naira per month for a movement from slow to fast broadband, 96% for a movement from fast to very fast broadband, and 90% for a movement from unreliable to reliable broadband. For an inclusion of the services of being able to: download high-Definition movies and TV shows (movie Rentals), make free video calls over the Internet (Videophone), and connect a laptop and other devices online wirelessly within a given distance around the home, consumers are on average willing to pay extra 89%, 87.5%, and 83% respectively. They are also, willing to pay 36.2% more in naira per month for the inclusion of the attribute of being able to designate some downloads as high-priority so that they travel through the Internet at relatively faster speed (priority). The consumer's increase in willingness to pay for broadband when there is an increase in the quality confirms their perception of the current quality of the broadband in the country as being poor (slow and unreliable). This also implies that their current feeling of the broadband as being expensive is relative to the broadband quality, which determines the level of satisfaction/utility they drive from the broadband. In other words, willingness to pay more for an increment in quality is not contradicting their initial assertion Table 1 Result for marginal willingness to pay for broadband. Dependent Variable: Log of Willingness to pay in Nigerian Naira per month (LnWTP(₦/month) Variables

Coef.

t-stat

Variables

Coef.

t-stat

Fast Very fast Reliability Home wireless

1.6634 0.964 0.9015 0.8332

8.1558 12.855 24.04 4.1486

Movie Rentals Tele-health Priority videophone

0.89 0.4948 0.362 0.875

9.33892 0.483296 1.1238 4.3174

8

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

Table 2 Result of WTP for broadband services given consumer characteristics. Dependent Variable: Log of Willingness to pay in Nigerian Naira per month (LnWTP(₦/month) Variables

Coef.

t-stat

Variables

Coef.

t-stat

Variables

Coef.

t-stat

Fast Aw(fast) Edu(fast) Urb(fast) Pub(fast) Poor(fast)

1.06 0.67 0.44 0.49 −0.71 −1.01

5.64 6.53 2.22 2.60 −3.45 −2.0

Very fast Aw(very fast) Edu(very fast) Urb(very fast) Pub(very fast) Poor(very fast)

0.99 0.59 0.39 0.57 −0.73 −0.95

5.28 17.89 3.44 1.54 −6.60 −12.1

Reliability Aw(Reliability) Edu(Reliability) Urb(Reliability) Pub(Reliability) Poor(Reliability)

0.67 0.56 0.4 0.26 −0.11 −0.66

5.77 11.24 4.95 0.94 −10.2 −2.38

that broadband services in the country are expensive, but just implying that they see it as expensive relative to its qualities and attributes. Relaxing the assumption of homogenous respondents, Table 2 below contains the result of the estimate of how the WTP for the inclusion of three main broadband attributes varies across individuals with heterogeneous socioeconomic and demographic characteristics. It shows that all the selected socio-economic and demographic characteristics of the individuals affect how they value improvement in broadband speed and therefore, their willingness to pay for such improvement. The coefficient of Aw (fast) is 0.67, and statistically significantly different from zero at 5% level of significance. This means that for an improvement from slow to fast broadband, respondents with deep knowledge of broadband are on average, willing to pay 67% more compared to those without broadband awareness. In addition, the coefficient Edu(fast) and Urb(fast) show that respondents with up to secondary education and those that live in urban areas are respectively willing to pay 44% and 49% higher compared to the uneducated and rural dwellers, for an improvement from slow to the fast broadband network. On the other hand, the coefficient of Pub(fast) and Poor(fast) are respectively −0.71 and −1.01 and are both statistically significantly less than zero at 5% level of significance. These suggest that on average, respondents in the public sector are willing to pay 71% less as extra amount for an improvement from slow to fast broadband than those in the private sector, and poor respondents are willing to pay 101% less than the rich ones as an extra amount for an improvement from low to fast broadband network. In fact, the poor respondents are only willing to pay 5.18% extra per month for an improvement from low to fast broadband service which is of course negligible. For a movement from fast to very fast broadband network, the results show that all the selected socio-economic and demographic characteristics of individuals affects how they value increase in broadband speed from fast to very fast and therefore, their willingness to pay for the very fast broadband network. All the coefficients except that of the place of residence are statistically significantly different from zero at 5% level of significance. The coefficient of Aw(very fast) is 0.594 which shows that on average, respondents with deep broadband awareness are willing to pay 59.4% more than those without broadband awareness as an extra amount for an improvement from fast to very fast broadband network. The coefficient of Edu(very fast) is 0.386, and this shows that on average, respondents with at least secondary education are willing to pay 38.6% more than those with less than secondary education as an extra amount for an improvement from fast to very fast broadband service. Furthermore, the coefficients of Pub(very fast) and Poor (very fast) are respectively −0.733 and −0.948, meaning that on average, respondents in the public sector and those that are poor are willing to pay 73.3% and 94.8% less than those in the private sector and the rich ones respectively for an improvement from fast to very fast broadband services. The third broadband attribute included in the estimation is reliability, and the consumers were asked how much extra they are willing to pay for an improvement in their broadband package from unreliable to reliable service. The results as contained in the Table 2 show that all the selected socio-economic and demographic characteristics of individuals influence how they value improvement in the reliability of broadband service and therefore, their willingness to pay for reliable broadband services. However, whereas all other coefficients are statistically significantly different from zero at 5% level of significance, that of the place of residence is not, showing that there is no significant difference in the willingness to pay for improvement in reliability of broadband between the urban and the rural dwellers in Nigeria. The coefficient of Aw(Reliable) is 0.56, and statistically significantly different from zero, which shows that on average, respondents that have deep broadband awareness are willing to pay 56% more than those without broadband awareness, for an improvement from unreliable to reliable broadband network. The coefficient of Edu(Reliable) is 0.4, indicating that on average, respondents with at least secondary education are willing to pay 40% more than to those with less than secondary education, as an extra amount for improvement from unreliable to reliable broadband service. The coefficient of Pub(Reliable) and Poor(Reliable) are respectively −0.11 and −0.66 meaning that on average, respondents in the public sector are willing to pay 11% less than those in the private sector, while poor respondents are willing to pay 66% less compared to the rich ones as an extra amount for an improvement from unreliable to reliable broadband.

4.4. Willingness to pay versus price and speed With the result of the current broadband being slow, the need of speed in the expansion of broadband demand, and the knowledge that increase in the speed of broadband will necessarily attract higher cost; the study went further to evaluate the actual amount 9

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

Fig. 5. Willingness to pay versus cost and speed of Internet.

consumers are willing to pay for different speed levels. The bar chart of Fig. 5 below shows that, though on average, willingness to pay decreases with increase in tariff rate (price), it increases with respect to increase in speed. In other words, for each tariff rate, increasing percentage of the consumers is willing to pay for higher broadband speed. However, the chart shows that as the monthly price of the broadband increases, the proportion of the consumers willing to demand broadband decreases and becomes insignificant if the broadband price goes beyond ₦10,000 per month, even at a speed of over 2Mbps. 4.5. Willingness to pay versus broadband reliability As a way forward, the study went further to analyze the consumers’ willingness to pay for increased broadband reliability. The results as contained on the split bar chart of Fig. 6 below show the percentage of the consumers that are willing to pay a given amount per month for an unlimited broadband bundle at different levels of reliability. On average, the graph depicts an exponential decrease in the general willingness to pay as the price per month of the broadband bundle increases and as the number of disruptions per period increases. This implies that at a given price, more consumers are willing to pay if the number of disruptions decreases and for a given number of disruptions, fewer consumers are willing to pay as the price increases. It is equally important to note that as monthly tariff rates go beyond 10,000 per month, not up to 3% of the sample is willing to pay even if the broadband is very reliable. Summarily, the result of the analyses shows that 79% of the respondents, still consider broadband in Nigerian as being expensive relative to its quality, which 58% and 84% see as slow and unreliable respectively. However, they are generally willing to increase their payment for broadband as its qualities and attributes increase with an upper price limit of ₦10,000 per month. This suggests that their rating of the current price of broadband as expensive is relative to its poor quality and not on absolute sense. In addition, their socioeconomic and demographic characteristics affect their willingness to pay: significantly and positively by their income, broadband awareness, and level of education, but insignificantly by their location. 5. Summary and recommendations 5.1. Summary Broadband network is now recognized globally as a very important proponent of economic growth and development. Hence, the Nigerian Government, Nigerian Telecommunications Commission (NCC), as well as the mobile network operators have been making efforts over the years to promote its use in the country. Despite these efforts, however, there is still low demand for broadband in

Fig. 6. Willingness to pay for broadband reliability. Source: Authors from the Survey data 10

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

Nigeria compared to the global economy and even other African Countries. This study, therefore, accessed consumer's perception of broadband cost and qualities, other factors affecting their demand for broadband, and their willingness to pay for an improvement in broadband qualities and attributes. It used a modified Savage and Waldman model and data from a nationwide survey conducted for Nigerian Communications Commissions in 2014, with a sample of 3300 respondents. The results show that 79%, of the respondents, still consider broadband in Nigerian as expensive relative to its quality, which 58% and 84% of them consider as slow and unreliable respectively. They are, however, generally willing to increase their current payment for improvement in the broadband qualities and attributes with an upper price limit of ₦10,000 per month, which show that they see the current broadband as expensive because of its low quality and poor attributes. In addition, their socioeconomic and demographic characteristics affect their willingness to pay: significantly and positively by their income, broadband awareness, and level of education, but insignificantly by their location. 5.2. Recommendations The result from this study shows that both the broadband suppliers and the government have a role to play in increasing consumers’ willingness to pay as well as demand for broadband in Nigeria. On the side of the suppliers, they should make an effort to increase the speed and ensure the reliability of the broadband supplied to the consumer. Undoubtedly, the required improvement in the quality of broadband supplied in the country will lead to increase in investment and thus, cost, which will ultimately be transferred to the consumers. However, one important finding from this study is that, although consumers see the current broadband price as expensive relative to their qualities and attributes, they are still willing to increase their payment for improved broadband qualities and attributes. Moreover, as a compliment, it may be necessary for the government to subsidize the use of broadband as the suppliers make efforts to improve on the qualities and attributes. The government should also develop initiatives to increase broadband literacy in schools, other public agencies, and institutions, and engage in community sensitization programs to increase broadband awareness, which is still very low in the country as found in this study. Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. It, however, used data from a survey funded by the Nigerian Communications Commission (NCC). References Atkinson, R. D., Correa, D. K., & Hedlund, J. A. (2008). Explaining international broadband leadership. Information Technology and Innovation Foundationhttp://www. itif.org/files/ExplainingBBLeadership.pdf, Accessed date: 28 June 2016. Duffy-Deno, K. (2000). Demand for high-speed access to the internet among internet households. Seattle: ICFC. http://www.icfc.ilstu.edu/icfcpapers00/duffy-deno.pdf/ accessed 01/06/2016, Accessed date: 1 June 2016. ERNST and YOUNG (2012). Mobile broadband Readiness: Key drivers to create positive environment. GSM Mobile Asia Expo June 2012. Retrieved on August 3, 2017, from www.gsma.com/spectrum/wpcontent/uploads/2012/06/Stephen_Lo_EY_Public_Policy_Forum_Mobile_Broadband_Readiness.pdf. Fairlie, R. (2004). Race and the digital divide. Contributions to economic analysis and policy. B. E. Press Journals, 3, 15–24 Article 15. FCC (2012). Measuring broadband in America july report: A report on consumer wireline broadband performance in the USFCC’s Office of Engineering, Technology, Consumer, and Governmental Affairs Bureau. available at: http://www.fcc.gov/measuring broadband-america/2012/july, Accessed date: 5 June 2016. Goolsbee, A. (2006). The value of broadband and the deadweight loss of taxing new technologies. Contributions to Economic Analysis and Policy, B. E. Press Journals, 5(1). Hassett, K. A., & Shapiro, R. J. (2009). Towards universal Broadband: Flexible broadband pricing and the digital divide. McDonough School of Business. ITU (2012). Regulating broadband prices. ITU universe of broadband portal. Jackson, M., Lookabaugh, T., Savage, S., Sicker, D. S., & Waldman, D. (2002). Broadband demand studyFinal Report. Telecommunications Research Group University of Colorado. Jagun, A., & Somolu, O. (2010). Broadband for Nigeria society. Markets & technology research. Jameel, A., Ehsan, N., Mirza, E., & Bhatti, M. W. (2012). Factors affecting Broadband proliferation in Pakistan and initiatives/measures taken by the Government of Pakistan. Procedia Technology, 1(2012), 250–257. http://dx.doi.org/10.1016/j.protcy.2012.02.051. Jon, H. (2012). More evidence of the economic importance of broadband. Warwickshire UK: Broadband Policy Consulting Ltd Wednesday, July 25, 2012. Kelly, T., & Rossotto, M. C. (2012). Broadband strategy handbook. Washington, D.C: The World Bank. MainOne Cable Nigeria LTD (2013). Broadband Nigeria–The next frontier. A paper presented at the Nigerian Communications Commission's stakeholder consultative forum for the year 2013 – 2017 strategic management plan. Manzoor, A. (2014). Investigating the factors affecting residential consumer adoption of broadband in India. http://dx.doi.org/10.1177/2158244014556998, sgo.sagepub. com SAGE Open October-December 2014. Mugeni, G. B., Wanyenbi, G. W., & Wafula, J. M. (2012). Evaluating factors affecting broadband readiness in Kenya: A pilot study. International Journal of Information and Communication Technology Research, 2(No. 6) June 2012. OECD (2008). Broadband and the economy. Ministerial background report DSTI/ICCP/IE (2007)3/FINAL (Paris). Ofcom (2012). UK fixed-line broadband performance, November 2011: The performance of fixed-line broadband delivered to UK residential consumers. The UK fixed broadband speeds November 2011: Research report. Ojobo, T. (2017). Nigeria Takes Broadband Internet Campaign to ITU Telecom World 2017. NCC Press Release https://www.ncc.gov.ng/stakeholder/media-public/publicnotices/231-press-release-nigeria-takes-broadband-internet-campaign-to-itu-telecom-world-2017. Okonji, E. (2017). Nigeria's broadband penetration hinged on mobile broadband. This Day newspaper of 27 July 2017. Accessed at https://www.thisdaylive.com/index. php/2017/07/27/nigerias-broadband-penetration-hinged-on-mobile-broadband/. Omigie, N. O., Zo, H., & Rho, J. J. (2015). User acceptance of mobile broadband in Nigeria. Information and Knowledge Management, 5(No.7)www.iiste.org. Parcu, P. L., Belloc, F., Cambini, C., Drouard, J., Manganelli, A., Nicita, A., et al. (2011). Broadband diffusion: Drivers and policies. Independent Regulation Group, Florence School of Regulation, Communication, and Media, European University Institute. Rappoport, P., Kridel, D., Taylor, L., Duffy-Deno, K., & Alleman, J. (2002). Residential demand for access to the internet. In G. Madden (Vol. Ed.), The international handbook of Telecommunications economics: Vol. II. Cheltenham: Edward Elgar Publishers. Rappoport, P., Taylor, L. D., & Kridel, D. J. (2003). Willingness to pay and the demand for broadband service. Mimeo.

11

Telecommunications Policy xxx (xxxx) xxx–xxx

N.E. Urama, O. Ogbu

Rosston, G., Savage, S. J., & Waldman, D. (2010). Household demand for broadband serviceFinal report to the Broadband.gov Task Force. Federal Communications Commission. Savage, S. J., & Waldman, D. M. (2009). Ability, location, and household demand for Internet bandwidth. International Journal of Industrial Organization, 27. Umezurike, C., Oludele, A., Kayuro, S. O., & Izang, A. A. (2015). Broadband internet penetration in Nigeria: A review. International Journal of Research Studies in Computer Science and Engineering (IJRSCSE), 2(Issue 1), 1–7 January 2015. Uzor, B., Jr. (2013). Lack of broadband policy leaves N347bn undersea cables redundant, in business day Wednesday. 16 January 2013 00:00. Varian, H. (2002). Demand for Bandwidth: Evidence from the INDEX project. Berkeley: Mimeo, University of California. Wareham, J., Levy, A., & Shi, W. (2004). Wireless diffusion and mobile computing: Implications for the digital divide. Telecommunications Policy, 28, 439–457.

12