Market Household Lighting Fuel Costs in Kenya This Market Intelligence Note presents empirical data from the 2005/06 Kenya Integrated Intelligence Household Budget Survey on the expenditure for lighting fuels and the adoption of solar Note home systems among Kenyan households. A key finding is that median expenditure on kerosene by Kenyan households that used this fuel for lighting was 2% of total income. Issue 2 Dec. 2012 Introduction1 According to the KIHBS, kerosene (in the survey also referred to as paraffin) was the dominant lighting fuel The Kenya Integrated Household Budget Survey (KIHBS) among Kenyan households (Table 1). Nearly three carried out by the Kenya National Bureau of Statistics quarters of all Kenyan households stated that their (KNBS) in 2005/06 offers unique insights into the main lighting fuel was kerosene, followed by electricity lighting fuel choices of a representative sample of and collected firewood (Table 1). Solar and dry cells Kenyan households. Moreover, by collecting data on a (torches) were less common, but still used by a number wide range of topics, it provides a solid quantitative of households as the main lighting fuel. The most basis for the analysis of many socio-economic common combination of lighting fuels for households questions. Specifically, the KIHBS dataset contains a that used more than one fuel was a combination of wealth of data related to the consumption of energy kerosene and dry cells. In cases where kerosene was sources in relation to lighting and cooking. It has been not the primary fuel, it was typically the secondary fuel used by a number of researchers to address questions (Lay et al., 2012). related to grid extension as opposed to off-grid electrification (Zeyringer et al., 2011), the demand for Table 1. Main sources of lighting, in percent (source: rural grid connections (Abdullah and Jeanty, 2009), and authors’ calculations based on KNBS, 2005/06) lighting fuel choice and the determinants of solar home Main Secondary system (SHS) adoption (Lay et al., 2012), among others. Lighting Fuel Lighting Fuel Fuel Type (%) (%) This Lighting Africa Market Intelligence Note is Collected specifically aimed at practitioners, policy-makers, and 6.4 12.3 firewood academics interested in the potential for and use of Purchased alternative lighting technologies (e.g., solar energy 0.2 1.1 firewood technologies) to displace the currently-prevailing Grass 0.2 0.2 lighting fuels (e.g., kerosene and firewood), which are Kerosene 73.5 20.8 commonly associated with substantial social, Electricity 16.4 2.0 environmental, and economic costs (Ekholm et al., Solar panels 1.4 1.4 Gas 0.2 0.9 2010). The Note builds upon a previous study by three Dry cell (torch) 1.4 48.3 of the authors, which contains additional information Candles 0.2 12.6 on the KIHBS methodology and dataset, as well as Biogas 0.2 0.1 research results on lighting fuel choice in general and Other 0.0 0.5 the determinants of SHS adoption in particular (Lay et Number of al., 2012). 12,990 4,479 observations Note: Table shows national data. 1 The authors of this note are Janosch Ondraczek, Jana Stöver, Jann Lay, and Arne Jacobson. Affiliations and acknowledgments are listed on the last page. © Lighting Global – December, 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 Kerosene Consumption lighting services. It should be stressed that when interpreting the following data, one must assume that Almost all households surveyed stated that they used not all electricity consumed was used for lighting kerosene in one way or another. Of those respondents purposes (especially in the case of higher income who used kerosene, the vast majority (81.2%) used it households). However, the survey shows that the primarily as a lighting fuel (Table A.1). Cooking was the primary use of electricity in the vast majority (95%) of second most important primary use of kerosene households is for lighting, suggesting that a major share (15.9%). of electricity is consumed for this purpose (Table A.3). Table A.1 suggests that the majority of kerosene Regarding absolute and relative expenditure on consumed was used for lighting, implying that most of electricity, the same pattern emerges as in the case of the stated kerosene expenditure must have been for kerosene. While absolute monthly expenditure meeting lighting needs. Because households reported increased by quartile (quartiles being defined over the monthly expenditures on kerosene and other fuels in whole population), readers should note the steep the KIHBS, absolute and relative expenditure levels on increase between the 3rd quartile (200 KSh/month) and kerosene can be derived by combining this information 4th quartile (500 KSh/month). This observation is even with stated annual household incomes and dividing the more important as the largest number of electricity survey sample into income quartiles.2 Assuming that users are in the 4th income quartile (1,153, or 72% of all most (if not all) kerosene was used for lighting, the data electricity users). Median monthly expenditure on suggest that the median Kenyan household spent electricity amounted to 332 KSh, or slightly less than 2% 156 KSh/month on kerosene for lighting, with absolute of the median household’s annual income. As seen monthly expenditure increasing by quartile. 3 The before for kerosene, the share of income spent on annual expenditure on kerosene decreased from electricity decreases from almost 3% to below 2% from almost 3% in the 1st quartile to a little over 1% in the 3rd the 1st quartile to the 4th quartile, but the overall quartile, with the median being slightly above 2% reduction is less dramatic than in the case of kerosene (Table A.2). (Table A.4). Electricity Use and Consumption Regarding electricity sources, the vast majority of households (88.7%) who use electricity reported Electricity was the second most important main lighting receiving power mainly from the Kenya Power & fuel (Table 1), but its use was mostly limited to Lighting Company (KPLC), Kenya’s national electricity households connected to the grid. Of course, electricity distribution company. 4 Solar panels are the second as a modern lighting fuel is not associated with some of most important source (9.2%) among those who use the negative socio-economic and environmental side- electricity at home, with the remaining 2.1% of effects mentioned in the introduction; therefore, data households indicating they receive power mainly from on electricity consumption should help to assess households’ willingness and ability to pay for modern 2 In line with the survey methodology, annual household expenditure is used 4 as a proxy for household income. Annual median income for the four Households were asked, “What is household main source of electricity?” if quartiles was 38,901 KSh (1st quartile), 74,610 KSh (2nd quartile), 118,511 KSh they reported using electricity (independent of source) or solar as their main (3rd quartile), and 236,639 KSh (4th quartile). or secondary lighting fuel. The range of possible sources they could then 3 The USD-KSh exchange rate in 2005/06 stood at 1:74 (i.e., KSh 156 identify were KPLC, community or personal generator, solar panels, car or corresponds to 2.11 current 2005/06 USD). motorcycle battery, or “other.” © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 diesel generators and car batteries.5 The median KPLC using a combination of kerosene as their main lighting customer spent 352 KSh/month on his or her electricity fuel and dry cells for torch lights (i.e., flashlights) as bill (slightly, but not significantly, higher than the their secondary lighting source (Lay et al., 2012). median presented in Table A.4). As discussed above, it According to the survey data, more than half of all is not possible to determine what fraction of this cost is households reported using dry cell batteries in the devoted to lighting services (KNBS, 2005/06). month prior to being interviewed, and nearly all households replied to the respective question Expenditure on Wood Fuel (Table A.7). Collected firewood was reported to be the third most Based on the information presented above, it can be important primary lighting fuel source and the fourth assumed that many of the dry cell batteries used by most important secondary lighting fuel source (Table 1). households served the purpose of powering a torch Firewood, both collected and purchased, was light. However, when interpreting the cost data for dry predominantly used for cooking (96.9% and 95%, cell batteries, readers need to once again consider that respectively). Very few households reported using it is unclear whether dry cell expenditure actually went wood primarily for lighting (Table A.5). This is not towards lighting services or whether the cells were surprising, as much of the lighting from firewood (as used for powering other electric devices, such as radios. seen in Table 1) was derived as a by-product of cooking With this caveat in mind, the data suggest that the with firewood. median Kenyan household spent 35 KSh/month on dry cells for all purposes, with absolute monthly No cost or expenditure data were reported for expenditure increasing from the 1st quartile to the 3rd collected firewood; therefore, this section presents quartile and then leveling off. Most striking, in the 1st some analysis based on purchased firewood, assuming income quartile the reported median expenditure on that findings in terms of absolute and relative spending dry cells was 0 KSh. Relative to annual income, the on purchased firewood can be used as approximations expenditure on dry cells initially increased from 0% in of households’ effort to collect firewood. This is an the 1st quartile to nearly 0.6% in the 2nd quartile, and imperfect assumption, but it does reflect the fact that then decreased to just 0.24% in the 4th quartile, with firewood collection involves opportunity costs the overall median expenditure being 0.29% associated with the time spent on collection. (Table A.8). Nonetheless, when interpreting the results presented in Table A.6, it is important to bear in mind that the Lighting Costs vast proportion of household firewood expenditure was for cooking purposes, not lighting. In a different section of the KIHBS survey, households were asked to report their monthly cost of lighting, Dry Cell Battery Consumption excluding the cost of electricity (which includes electricity from solar panels). As can be seen from According to the data presented in Table 1, dry cell Table A.9, these data allow for the presentation of batteries were the most frequently named secondary households’ median overall expenditure on lighting for lighting fuel source (48.3%), with most households both their stated main and secondary lighting fuel.6 5 6 The limited importance of car batteries reported by households contrasts It should be noted that the expenditure reported here is for all lighting with earlier research that found a large number of households using such fuels and the table therefore does not allow one to disaggregate batteries for their off-grid electricity needs (e.g., Jacobson, 2007). expenditure by main, secondary, and other lighting fuels. © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 Table A.9 suggests that households using kerosene as The total monthly cost of lighting (excluding electricity) their main lighting fuel had a median expenditure of for the median household using kerosene as its main 150 KSh/month for lighting (most of which was lighting fuel increased from 100 KSh/month for the 1st probably spent on kerosene), whereas electricity users quartile to 250 KSh/month for the 4th quartile. spent an additional 50 KSh on other lighting fuels. Conversely, for households with kerosene as their Households using collected firewood as the main secondary lighting fuel, no such increase could be lighting fuel spent a median 100 KSh/month on lighting observed (Table A.10). This suggests that households (with a high probability of this being kerosene). using kerosene only as their secondary lighting fuel Interestingly, owners of solar home systems (SHS) (and hence some other fuel as their main lighting fuel) seemed to need substantial amounts of additional generally spent less on lighting overall (excluding lighting fuels to meet their overall lighting demand, as electricity). Likewise, households using dry cells as their they reported spending a median 120 KSh/month on secondary lighting fuel reported median total costs as other lighting fuels. rising from 110 KSh/month for the 1st quartile to 250 KSh/month for the 4th quartile (Table A.11).7 Households using dry cells as their secondary lighting fuel reported a median monthly expenditure of 161 KSh The same pattern of absolute spending on lighting for all lighting fuels (Table A.9). Based on the evidence (excluding electricity) increasing by quartile can also be presented before, this expenditure was likely to be split observed in the case of collected firewood as the primarily between kerosene (as the most common secondary lighting fuel, where the median monthly main fuel source) and dry cells (as the most common expenditure steadily increased from 80 KSh/month for secondary fuel source). However, it is unfortunately not the 1st quartile to 180 KSh/month for the 4th quartile possible to determine from the data which fuel source (2nd quartile: 118 KSh/month, 3rd quartile: accounted for which proportion. Readers should note 174 KSh/month). Interestingly, median expenditure if the close match between the expenditure levels candles are used as the secondary fuel remained fairly reported in Table A.9 and the expenditure levels stable between 50 KSh/month and 100 KSh/month presented in earlier tables. across quartiles (1st and 4th quartiles: 50 KSh/month; 2nd and 3rd quartiles: 100 KSh/month). The overall number The low reported median monthly cost for households of observations is relatively small (N = 428), however, using kerosene as their secondary lighting fuel and most of these observations correspond to the 3rd (60 KSh/month) may stem from these households using and 4th quartiles. collected firewood as their main source. Households using collected firewood as their main source of lighting Conclusions spent little on other lighting fuels, and in many cases these expenditures were for kerosene as their “backup” Kerosene is the single most important lighting fuel used fuel. On the other hand, households using candles as in Kenya, as evidenced by a median monthly their secondary fuel (median expenditure on lighting of expenditure on kerosene of 156 KSh according to the 69 KSh/month) predominantly relied on electricity 2005/06 KIHBS survey. The second most important (59.2%) and kerosene (39.6%) as their main fuels, of lighting fuel, electricity, is a modern fuel and which only the latter should be represented in the stated total cost of lighting. 7 Readers should note that Table A.10 and Table A.11 present the total cost of lighting, whereas the corresponding numbers in Table A.2 and Table A.8 show the stated expenditure on kerosene and dry cells, respectively. Therefore, the tables cannot be compared directly. © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 households using it spent a median 332 KSh/month to 6. Zeyringer, M., Morawetz, U., Pachauri, S., Schmid, use it. However, households in the lower income E., Schmidt, J., (2011), Stand-alone vs. grid quartiles were typically not connected to the grid. extension for electrification in Kenya - Lighting from wood is usually a by-product of cooking Development of a spatial explicit energy system with firewood. Therefore, expenditure on wood model, Proceedings of the 34th IAEE International specifically for lighting was likely negligible. Dry cells Conference, Stockholm. were the fourth most important secondary lighting fuel, but median expenditure on dry cells was relatively low (35 KSh/month). In contrast, owners of solar home Report Authors: Janosch Ondraczeka,*, Jana Stöverb, systems (SHS) reported spending a median Jann Layc, Arne Jacobsond 120 KSh/month on alternative lighting fuels—primarily kerosene—to meet their overall lighting needs, Affiliations: a University of Hamburg, Research Unit compared to only 50 KSh/month spent on alternative Sustainability and Global Change (FNU), Germany, lighting fuels by other electricity users. This suggests *corresponding author: janosch.ondraczek@zmaw.de; b Hamburg Institute of International Economics (HWWI) that the SHS in place did not enable households to and University of Hamburg, Germany; c GIGA German meet their total lighting needs without additional Institute of Global and Area Studies, Hamburg, and lighting fuels. University of Göttingen, Germany; d Schatz Energy Research Center, Humboldt State University, USA. References Acknowledgement: The authors would like to thank 1. Abdullah, S., Jeanty, P.W. (2009), Demand for the Kenya National Bureau of Statistics for providing Electricity Connection in Rural Areas: The Case of the KIHBS dataset. Kenya, Working Paper 26/09, Department of Economics, University of Bath. 2. Ekholm, T., Krey, V., Pachauri, S., Riahi, K. (2010), Determinants of household energy consumption in India, Energy Policy 38, 5696-5707. 3. Jacobson, A. (2007), “Connective Power: Solar Electrification and Social Change in Kenya," World Development, v35, n1, pp. 144-162. 4. KNBS (2005/06), Kenya Integrated Household Budget Survey 2005-2006, Version 1.0 of the KNBS dataset, Kenya National Bureau of Statistics, Nairobi. 5. Lay, J., Ondraczek, J., Stöver, J. (2012), Renewables in the Energy Transition: Evidence on Solar Home Systems and Lighting-Fuel Choice in Kenya, GIGA Working Paper No 198, July 2012, Hamburg. © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 Appendix A. Table A.1. Primary use of kerosene (source: authors’ calculations based on KNBS, 2005/06) Purpose of use Frequency Percent Boiling 206 1.9 Heating 83 0.8 Cooking 1,747 15.9 Lighting 8,947 81.2 Other 42 0.4 Number of observations 11,025 100.0 Note: Table shows national data. Table A.2. Median expenditure on kerosene by quartile, absolute, and as a percentage of annual income (source: author's calculations based on KNBS, 2005/06) Absolute Monthly Expenditure Percent of Annual Income Income Quartile (KSh)* Spent on Fuel 1 100 2.91 2 150 2.30 3 200 1.96 4 250 1.26 Total 156 2.01 Note: Table shows national data. * Blank data entries for kerosene expenditure data for households that reported they were not using kerosene was set to 0 KSh. This reduces median monthly expenditure on kerosene from 160 KSh to the 156 KSh presented here (a reduction of 2.5%). Table A.3. Primary use of electricity (source: authors’ calculations based on KNBS, 2005/06) Purpose of use Frequency Percent Boiling 36 1.8 Cooking 49 2.4 Lighting 1,927 95.0 Other 17 0.8 Number of observations 2,029 100.0 Note: Table show national data. © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 Table A.4. Median expenditure on electricity by quartile, absolute, and as a percentage of annual income (source: author's calculations based on KNBS, 2005/06) Absolute Monthly Expenditure Percent of Annual Income Income Quartile (KSh)* Spent on Fuel 1 100 2.74 2 150 2.15 3 200 1.95 4 500 1.82 Total 332 1.93 Note: Table shows national data. * Blank data entries for electricity expenditure data for households that reported they were not using electricity was set to 0 KSh. This reduces the median monthly expenditure from 350 KSh to the 332 KSh presented here (a reduction of 5.1%). Table A.5. Primary use of collected and purchased firewood (source: authors’ calculations based on KNBS, 2005/06) Collected Firewood Purchased Firewood Purpose of use Frequency Percent Frequency Percent Boiling 109 1.5 64 3.4 Cooking 61 0.8 18 0.9 Lighting 7,280 96.9 1,817 95.0 Other 63 0.8 10 0.5 Number of observations 7,516 100.0 1,912 100.0 Note: Table shows national data. Table A.6. Median expenditure on purchased firewood by quartile, absolute, and as a percentage of annual income (source: author's calculations based on KNBS, 2005/06) Absolute Monthly Expenditure Percent of Annual Income Income Quartile (KSh)* Spent on Fuel 1 50 1.53 2 100 1.40 3 150 1.55 4 250 1.21 Total 120 1.38 Note: Table shows national data. * Blank data entries for firewood expenditure data for households that reported they were not using purchased firewood was set to 0 KSh. This reduces the median monthly expenditure from 200 KSh to the 120 KSh presented here (a reduction of 40%). © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 Table A.7. Use of torch batteries (dry cells) (source: authors’ calculations based on KNBS, 2005/06) Batteries used Frequency Percent Yes 7,187 55.3 No 5,802 44.7 Number of observations 12,989 100.0 Note: Table shows national data. Table A.8. Median expenditure on dry cells by quartile, absolute, and as a percentage of annual income (source: author's calculations based on KNBS, 2005/06) Absolute Monthly Expenditure Percent of Annual Income Income Quartile (KSh)* Spent on Fuel 1 0 0.00 2 35 0.57 3 50 0.51 4 50 0.24 Total 35 0.29 Note: Table shows national data. * Blank data entries for dry cell expenditure data for households that reported they were not using dry cells was set to 0 KSh. This reduces the median monthly expenditure from 80 KSh to the 35 KSh presented here (a reduction of 56%). Table A.9. Median total cost of lighting (excluding electricity) by fuel source (source: authors’ calculations based on KNBS, 2005/06) Main Lighting Fuel Secondary Lighting Fuel Cost Cost Fuel Type (KSh/month) Number of Responses (KSh/month) Number of Responses Collected firewood 100 258 104 532 Purchased firewood 135 24 135 46 Grass 150 19 150 7 Kerosene 150 9,437 60 736 Electricity 50 1,024 150 49 Solar panels 120 121 200 59 Gas 175 26 180 33 Dry cell (torch) 100 70 161 2,086 Candles 100 27 69 428 Biogas n.a. n.a. 104 1 Other 240 7 230 18 Total 150 11,013 140 3,995 Note: Table shows national data © Lighting Africa – December 2012 Household Lighting Fuel Costs in Kenya Issue 2 December, 2012 Table A.10. Median total cost of lighting (excluding electricity) if kerosene is used as the main or secondary fuel, by quartile (source: author’s calculations based on KNBS, 2005/06) Main Lighting Fuel Secondary Lighting Fuel Income Quartile (KSh/month) (KSh/month) 1 100 69 2 150 50 3 200 60 4 250 80 Total 150 60 Number of Observations 9,434 735 Note: Table shows national data Table A.11. Median total cost of lighting (excluding electricity) if dry cells are used as the secondary fuel, by quartile (source: author’s calculations based on KNBS, 2005/06) Cost of Lighting Income Quartile (KSh/month) Number of Responses 1 110 359 2 150 567 3 200 644 4 250 515 Total 161 2,085 Note: Table shows national data. © Lighting Africa – December 2012