Introduction

One of the factors driving the ongoing Power crisis in India is the surge in demand for electricity by households to run appliances that provide relief from heat. A better forecasting and managing of demand would have possibly reduced the extent of the crisis. Both short term forecasting and long-term projections of residential electricity consumption require data on appliance ownership, efficiency levels, and usage. However, India does not conduct a periodic national residential energy consumption survey like some countries. The last extensive official data on appliance ownership is from the NSSO's Household Consumer Expenditure survey in 2012. Several surveys with different objectives and sample sizes have been conducted by various organisations which include questions on appliance ownership among other things. These can be used to gather data on appliance ownership and then to estimate residential electricity consumption. The Consumer Pyramids Household Survey (CPHS^[1]) conducted by the Centre for Monitoring Indian Economy (CMIE^[2]) is one such survey.

CPHS is a longitudinal survey conducted three times a year and has covered over 200,000 households since 2014.  One round of the survey covers four consecutive months in a calendar year, and is called a wave. CPHS covers a gamut of aspects related to living standards of Indian households including income, ownership of assets, access to different services, and expenditure under different categories. As per CMIE, CPHS is representative at national, state and urban-rural levels. A number of researchers have used the data to examine income equality, access to finance, unemployment and other issues.  Some concerns have also been raised about the sampling methodology and the survey process. We will briefly discuss them later in the context of our findings.

We have used data from wave 1 (first round of 2014) to 22 (first round of 2021) to examine ownership trends of electrical appliances like television, refrigerator, air cooler, air conditioner, washing machine, and computer. We have also analysed the data on ownership of 2-wheelers, 4-wheelers, and inverter-generators. We have three broad observations which we discuss below. We use the relevant weights provided by CPHS to extrapolate the outcomes from sample to the population^[3].

Appliance ownership seems to be on the higher side

We use the data from the last round of 2019 (Wave 18) to comment on the appliance ownership before the pandemic. The pandemic has likely affected both, the actual ownership patterns and the survey process. Hence, there are some uncertainties in the data accuracy during the pandemic time which we will discuss in the next section.  As seen in Figure 1, about 95% of the households owned a television, 72% owned a two-wheeler, and 60% owned a refrigerator in 2019.

Figure 1: Appliance ownership in 2019

In order to examine the data in a more granular way, we split the sample into urban and rural households and into quintiles (5 groups of equal number of households) based on income. We observe that most of the top urban quintile (highest income) households own a TV, 2-wheeler, and a refrigerator followed by a washing machine with their rural counterparts owning fewer appliances such as refrigerators and washing machines. About 42% of the urban households in the topmost quintile own an air-conditioner. In the bottom rural quintile, 86% of the households own a TV, 47% own a 2-wheeler, and 35% a refrigerator.

Figure 2: Appliance ownership in the highest and lowest quintile in rural/urban region in 2019

We compared these results with the National Family Health Survey (NFHS)^[4], a multiple-round survey conducted for the Ministry of Health and Family Welfare (MoHFW) by the International Institute for Population Sciences, Mumbai. NFHS' last round (NFHS-5) of survey was conducted over 2019-21 and the one before that in 2015-16 (NFHS 4). We compared the data with the corresponding CPHS waves. About 572,000 households were surveyed across India in NFHS 4 and 636,000 households in NFHS 5. We observe that CPHS reports significantly higher ownership of appliances compared to NFHS for most of the appliances.  Higher ownership of appliances in CPHS compared to NFHS is indicative of CPHS' bias towards higher income households as has been pointed out by other researchers^[5].

Figure 3: Comparison of CPHS with National Family Health Survey (NFHS)

Appliance ownership has increased significantly between 2014 and 2019

As CPHS is a longitudinal survey, it is interesting to observe the change in ownership patterns of such a large sample of households even after considering the possibility of bias towards high income households. We see three distinct phases of change in appliance ownership from 2014 to 2019 (see Figure 4). There is a sudden increase in ownership of all the appliances over the three rounds of the year of 2014. This is followed by almost a plateau till 2016. Starting 2017, there has been a strong, consistent growth in ownership with the share of households owning 2-wheelers increasing from 50% to 72%, refrigerators increasing from 42% to 60%, and coolers increasing from 29% to 47%.

Figure 4: Appliance ownership trends between 2014 and 2019

This pattern is observed more or less in both urban and rural households as well as those in different quintile groups (see Figure 5).

Figure 5: Appliance ownership trends between 2014 and 2019 for urban/rural

This observed change in ownership patterns is difficult to explain. The only significant event in 2014 was the national general election. There may be a possibility of some appliances being distributed free during the elections but that hardly explains the significant jump in the ownership. The second phase of increase in ownership post demonetization is complex.  CPHS shows significant growth in ownership of all appliances across all income groups. This is also exhibited in the trends of household incomes over the same time (Figure 6). However, this is contrary to the general slowdown of consumer demand^[6] post 2016 as indicated by the drop in growth rates of GDP and private final consumption expenditure (PFCE). This needs to be examined further.

Figure 6: Trends in median income of top and bottom quintile from 2014 to 2019

Appliance ownership fluctuates over time

CPHS data shows that appliance ownership fluctuates over time. This is particularly observed during the two macro shocks observed in India, demonetization and pandemic. The household appliance ownership of several appliances dropped immediately following the event and then reached back to the pre-event levels in a short time. This is observed across all income groups. This is much pronounced in some states like Andhra Pradesh (Figure 7). One interpretation of this behaviour can be, that households sold their appliances in times of economic distress and then bought them back in a short time after conditions improved. This is also observed in higher income group households. However, this is contrary to the expected behaviour. Appliances like refrigerators, TVs, coolers once purchased are generally used for at least 10-15 years before they are replaced. Their price depreciates fast and it does not make economic sense to sell them during economic stress and buy later. Also, there should be some buyers for these appliances and since it is a nationally representative survey, it should mean that the overall ownership remains same. This also needs further examination.

Figure 7: Appliance ownership trends in Andhra Pradesh

Conclusion

Our analysis indicates that the CPHS is biased towards high income households, which has also been pointed out by several researchers^{[7] [8] [9]} recently in the context of other parameters such as literacy rate, access to finance.  The changes in ownership patterns of appliances over the different waves of CPHS present interesting unexpected trends even after considering the possibility of bias towards high income households. Sudden phases of increase in ownership of appliances and appliance ownership fluctuation over time are contrary to expected behaviour. This needs to be examined further before using it to make any inferences about appliance ownership and its impact on residential electricity consumption in India.

Endnotes