- Background The largest nationally representative population study by Indian Council of Medical Research-India Diabetes (ICMR-INDIAB) reported that an estimated 101 million people in India have diabetes and 136 million have prediabetes (approximately 1 in 10). The experts claim that the high number of prediabetics is more worrisome because it suggests that the diabetes “epidemic” is far from over, and therefore the situation is expected to worsen in the future. A pan India study (2020) about the cost of management of diabetes, reports that approx. 17% of the expenditure of the household was on health care; the average expenditure on diabetes medicines in rural areas is ₹ 1,072.28, while that in urban areas is ₹ 1,135.61. In a country where the per capita gross national disposable income in 2020 was ₹ 12,708, spending around 10% of that income on diabetic medicines seems exorbitant. As per the Union Budget (2023), the percentage spent on health care as a percentage of GDP is roughly 2.7%. Bangladesh, which like India is a developing country, spends approx. 5.6% of its Gross Domestic Product (GDP) on healthcare, i.e. twice that of India. Therefore, it can be reasonably assumed that the current governmental interventions are highly inadequate.
Researchers have not been able to identify the absolute causes of Type-1 and Type-2 diabetes because this disease can occur from a combination of factors. However, the condition of prediabetes is curable to a greater extent by making lifestyle changes. One such change is replacing foods with high glycemic index (GI), such as rice, with foods having a low GI (e.g. millets, such as Ragi, jowar, bajra). Although a few FMCG companies, such as Tata Consumer Products and MTR sell Ragi based products, Ragi consumption is low in comparison to jowar and bajra (Figure 2).
2 Scope
This article aims to promote Ragi consumption using inferences from a linear regression-based demand estimation model. We hypothesize that diabetics do not consume much Ragi products, because the product packaging does not explicitly state that it is “diabetic friendly” or “made for diabetics”. If Ragi is marketed as a “food for diabetics”, rather than a generic healthy food, then it will stand out amongst the plethora other “Ready to Eat” healthy snacks on the same aisle of the supermarket. The Startups companies may use the insights provided in the article to better position the Ragi products to increase its consumption.
- Reasons for less consumption of millets
The International Crop Research Institute for the Semi-Arid Tropics (ICRISAT) conducted a consumer survey in which 15,500 one to one interview were recorded across 7 states in urban shopping centers. The survey indicated that 91% of the respondents were reasonably health conscious. However, only 40% were aware that millets were healthy. Therefore, it is inferred that the awareness of the health benefits of millets is limited. 30% of the respondents said that the single largest reason for eating millets was due to a health problem (Figure 9), followed by 15% of the respondents, who consumed them to lose weight. 40% of the respondents said that they did not eat millets because it was not traditionally eaten (Figure 10) by their families. Most respondents (50.7%) said that the primary source of food nutrition related information is social media followed by television advertisements and word of mouth. Interestingly, the taste of millet products was not a significant factor in the respondents’ decision to consume millet. But, the survey report emphasized the need for tasty millet products (Figure 11). Regarding the main sources of information about health, the majority respondents (51%) said that they rely on “social media” for health-related information. Only 12.5% of the participants commented that millets are a good substitute for people with diabetes who cannot consume enough rice, because of its high glycemic index.
The ICRISAT survey, therefore, suggests that the marketing campaigns about millet products can have a ripple effect on consumption if they target the decision maker who buys groceries in the household. Also, it can be inferred from the survey responses that a specific health problem is an influential driver for people to eat millets, rather than consuming it just because millets being healthy.
3 Linear Regression
3.1 Inputs for Linear Regression Modelling
Linear regression analysis is used to predict the value of a dependent variable based on the values of other variable(s) called independent variable(s). Because of the constraints in availability of several data types across common timeframes, only the data between 2000 and 2019 were selected. More information about Linear Regression can be read on this website. Table 1 includes the linear regression variables, sources of data, and extrapolations/ assumptions (if any) made to perform the linear regression.
Table 1: Inputs for creating a Linear Regression model
# | Variable | Data Source | Extrapolation method/ Assumptions regarding data |
1 | Dependent Variable |
|
|
1.1 | Total Ragi demand (sum of Rural and Urban consumption) (in Tonnes)
| Research paper | Since the Ragi consumption data was not available, the total Ragi demand was extrapolated using multiple sources. The urban and rural consumption was provided in the research paper for the years 1993-94, 1999-00, 2004-05 and 2011-12 only based on the 50th, 55th, 61st and 68th surveys of the National Sample Surveys Organization (NSSO). The data points between 2000 and 2012 were extrapolated by averaging and distributing the growth rate between each 5-year time period. To find the data points between 2012 and 2019, an interview of an industry expert in the article was used. The expert said that the consumption of Ragi increased 15% between 2012 and 2022. So, the growth rate of 15% was evenly distributed between 2012 and 2019 to find 7 data points. Thus, in total 19 data points of data about total Ragi consumption. |
2 | Independent Variables | Data Source | Extrapolation method/ Assumptions regarding data |
2.1 | Adults with diabetes (age 15 – 64 yrs.) | World Bank dataset and Research Paper | The International Diabetes Federation (IDF) states that India had approx. 77 million of the total Indian population suffering from diabetes in 2019. The data about the count of Indian population between 15 to 64 years age is sourced from the World Bank dataset. It is assumed that the proportion of people suffering from diabetes in the 15 to 64 years age group is the same as that of proportion of diabetics in total population. The research paper states that the rate of increase of diabetes amongst Indians is between 3 to 5 % per year. The growth rate between 3 and 5% was randomized to get 19 random growth rate values between 3 and 5%. These growth rates were used to estimate the number of adults (15 to 64 years age) with diabetes from 2000 to 2019. |
2.2 | Rice Consumption (in Metric Tonnes) | Indexmundi Dataset | N/A |
2.3 | Wheat Consumption (in Metric Tonnes) | Indexmundi Dataset | N/A |
2.4 | Gross National Disposable Income per capita (INR) | Ministry of Statistics and Program Implementation via CEIC database | N/A |
2.5 | Minimum Support Prices (Rs./Quintal) | Directorate of Economics and Statistics, Department of Agriculture and Farmers Welfare via CEIC database | N/A |
2.6 | Average Prices of Ragi in Domestic Market Rs. / Quintal | CMIE Commodities Dataset | N/A |
2.7 | Average Prices of Rice in Domestic Market Rs. / Quintal | N/A | |
2.8 | Average Prices of Wheat in Domestic Market Rs. / Quintal | N/A | |
2.9 | Average Prices of Sorghum Millet (Jowar) in Domestic Market Rs. / Quintal | N/A | |
3.0 | Average Prices of Pearl Millet (Bajra) in Domestic Market Rs. / Quintal | N/A | |
3.1 | Average Prices of Maize in Domestic Market Rs. / Quintal | N/A |
3.2 Results of Linear Regression (Part 1)
- Case 1
Figure 3 shows the regression results using Total Ragi Demand as Dependent Variable and other columns as Independent Variables. The independent variables in blue color are significant (p-value < 0.05). The variable Adult with Diabetes is significant with a negative coefficient. Thus, it can be understood based on the regression model that the count of diabetics is negatively correlated with the consumption of Ragi. In short, the diabetics are not consuming enough Ragi, and therefore the increase in their number is negatively correlated with Ragi consumption. Interestingly, a few variables, such as Rice and Wheat consumption, which were expected to be significant, did not appear to be significant in the regression result. One reason could be the issue of multicollinearity in the dataset within the independent variables. Therefore, in Cases 2, 3 and 4, only a few independent variables were considered to check whether they could be significant if the issue of multicollinearity is diminished.
- Case 2
Figure 4 shows the regression results using Total Ragi Demand as Dependent Variable and only Rice & Wheat consumption as independent variables. Rice consumption is a significant variable (p < 0.05) affecting Ragi demand. Since it is negatively correlated, it can be said that based on the regression model, if the rice consumption increases, then ragi consumption decreases, by a factor of 0.03 (ceteris paribus). - Case 3
Figure 5 shows the regression using Adults with Diabetes as Dependent Variable and only Rice & Wheat consumption as independent variables. Both Rice and Wheat consumption are significant (p < 0.05) variables in affecting the count of diabetics in India. Therefore, it can be reasonably concluded that based on the regression model, as the Rice and Wheat consumption increases, the number of people with diabetes in India also increases. - Case 4
Figure 6 shows the regression results using Adults with Diabetes as Dependent Variable and only Rural Ragi Consumption and Urban Ragi Consumption as independent variables. The variable Rural Ragi consumption is significant (p < 0.05) with a highly negative coefficient. From the regression result, it can be concluded that as the Rural Ragi consumption decreases, the count of diabetics in India increases by a large factor. Therefore, based on the regression model, the Rural Ragi consumption must increase to have the maximum impact at reducing the count of diabetics in India. The results also indicate that despite the benefits of Ragi for people with diabetes, it is still unknown within its segment of cereals/ healthy/ snacks/ food grains. But, if it is marketed as a “food for diabetics”, then it will differentiate itself amongst other products. Our secondary market research via online grocery websites (as of June 2023), revealed that only Britannia Nutri Choice biscuits are explicitly marketed as “Diabetic Friendly”. Our team did not find any other product using the terms “diabetic food”, “good for diabetics” etc. on the packaging of Ragi products. Therefore, it can be reasonably concluded that a gap exists in the market for Ragi foods targeted to “diabetics”.
- Results of Linear Regression (Part 2)
- Analysis of Coefficients in Regression Results
- Income Elasticity
According to the Regression Results in Figure 3, the coefficient of independent variable “Gross National Disposable Income Per Capita” is 0.02. This means that the income elasticity of Ragi is positive. It indicates that Ragi is a normal good, i.e., people buy more Ragi as their disposable income increases.
According to a recent study (2022), the rate of reduction in consumption of Ragi was comparatively higher in the case of rural households than urban. A possible reason could be that Ragi is not consumed as a staple food in most of India. Ragi products are generally marketed as healthy superfoods, which are sold at a premium in urban areas where consumption is largely increasing. Therefore, if Ragi products are promoted in urban areas as a “food for diabetics” rather than as a generic healthy product, they are more likely to be sold. - Cross price Elasticity:
According to the Regression Results in Figure 3, the coefficient of independent variable “Average Prices of Pearl Millet (Bajra) in Domestic Market (Rs./ Quintal)” is 0.63. This means that the cross-price elasticity of Bajra is positive. Bajra is a natural substitute of Ragi as they have some similarities in terms of their nutritional content and culinary usage. The cross price elasticity also depends on the availability of grains as the availability may vary depending on geographies and regional/ local preferences. In some regions or markets, Ragi may be less accessible or more expensive compared to Bajra and hence Bajra can act as a substitute for Ragi. - Minimum Support Price (MSP) and Own Price Elasticity Minimum Support Price (MSP) and Own Price Elasticity may be considered important in the demand equation of a commodity, such as Ragi. However, we have ignored these factors because these variables are not significant (p>0.05) (Figure 3). The same has happened due to multi-collinearity between variables which is a major drawback of linear regression. Our primary objective for this article is to evaluate whether Ragi can be positioned and marketed as a “food for diabetics” from the point of view of an Startup company. The price elasticity of Ragi is not relevant for this segment, i.e. the urban segment that has access to a supermarket of an Startup company. Although we have briefly discussed in Case 4 on the how increasing Rural Ragi consumption can have a greater positive impact on reducing the count of diabetics in India, our main hypothesis (see Scope) is towards the urban segment, which is serviced by an Startup company as it is more profitable than the rural segments.
Inferences from Regression ResultsTo capitalize on the gap (see Case 4) identified in the marketing of Ragi food products, we propose that the Startup companies use the inferences derived from the regression results.Before re-positioning an existing Ragi product or before releasing a new Ragi product as a “food for diabetics”, the Startup company should make efforts towards creating such conditions that the coefficient of “Adults with Diabetes” (Figure 3) in the target urban city is positive rather than negative. A positive coefficient would mean that the diabetics have begun to include Ragi as a staple food in their diets. Section 2.1 provides insights from the ICRISAT survey about low awareness of the benefits of consuming Ragi. A Startup company may create a “Ragi for Diabetes” campaign as a pre-release marketing initiative that projects the benefits of Ragi as a “substitute for rice” based on the high positive correlation of rice consumption with diabetes (discussed in Figure 5).
Moreover, Ragi has many nutritional benefits (Figure 8) over other millets (Jowar, Bajra etc.). Ragi has the highest fiber content, Calcium and Potassium content in comparison to other millets. It has less fat than Bajra and the lowest GI, i.e. it does not cause spikes in blood sugar levels.Government interventions along with global support can increase the reach of Ragi food products as a “food for diabetics” in rural India. The International Year of Millets (IYoM 2023) is one such global initiative. IYOM was recognized by the United Nations General Assembly at its 75th session in March 2021. Millets are tolerant of climatic fluctuations and may be grown on arid terrain with few inputs. As a result, they are the perfect way for nations to boost national independence and lessen their dependence on imported cereal grains. Along a similar line of thought, we hope this article nudges the Startup companies in promoting domestic Ragi products over foreign oat cereal brands (e.g. Dr. Oetker, Quaker Oats). These brands are marketed heavily by conglomerates, such as PepsiCo and Kelloggs’, although the bioavailability of minerals in oats is less than Ragi.