The Effect of Different Occupations on Health Outcomes in India: A Gender-Based Analysis, Summaries of Economics

Download The Effect of Different Occupations on Health Outcomes in India: A Gender-Based Analysis and more Summaries Economics in PDF only on Docsity!

THE EFFECT OF DIFFERENT OCCUPATIONS IN EXPLAINING

HEALTH OUTCOMES DIFFERS BY GENDER OVER TIME IN INDIA

Anulya P^1

1. Department of Economics, Ashoka University

ABSTRACT

India's vast and diverse workforce faces a multitude of health challenges, including obesity,

overweight, underweight, and diabetes. This study, based on data from the National Family

Health Survey (NFHS-3 and NFHS-4), investigates the prevalence of these health issues and

how they vary between men and women. The paper employs a linear probability model to

analyze the trends in obesity and diabetes over time. Our findings reveal that individuals

engaged in white-collar occupations are at a higher risk of obesity. In contrast, those in blue-

collar jobs, including agriculture, exhibit a spectrum from underweight to overweight over

different time periods. Women tend to fluctuate between underweight and overweight,

relatively protected from obesity. These results underscore disparities in obesity, underweight,

and overweight between genders. Additionally, our analysis indicates that men are more

susceptible to diabetes than women, with both genders experiencing an increased prevalence

of diabetes over time. Notably, individuals in white-collar jobs are more prone to diabetes

compared to their counterparts in blue-collar positions. This study provides essential insights

into the distribution of health issues across different occupational sectors in India, emphasizing

the need for tailored interventions to address these disparities and improve overall public

health.

JEL Classifications : I12, I18, I19, J

Keywords: Structural change of occupation, Health, Obesity, Diabetes, Overweight,

Underweight, Gender, Time, Skills

1. INTRODUCTION

Occupation is an intrinsic factor of subsistence. India has a large number of working

populations involved in diverse occupations. If broadly classified, there would be two type

jobs: white-collar and blue-collar jobs. Blue-collar jobs are those that involve a greater degree

of physically taxing or manual labour.^1 White-collar jobs, on the other hand, typically work in

(^1) Charles N. Weaver, “Job preferences of white collar and blue-collar workers,” Academy of Management Journal 18, no. 1 (1975): 167-175.

Similarly, in Barlin & Mercan (2016)^4 , examined the effect of working hours of obesity by

occupation groups in the US. The study found that workers in some occupation groups have

less likelihood for obesity. Identified that six groups of occupations lower the risk of obesity,

but they were unable to find the mechanism behind the reduced risk in the groups. The

assumption was due to the different nature of occupations and profile of the members of

occupation groups. Because some groups of occupation involve high physical activity (comes

under blue collar jobs) whereas some require higher education (comes under white collar jobs).

Likewise, Farinelli et.al (2015)^5 also observed that the patterns of overweight and obesity

among occupations differs by gender. It was rendering an indication for health lifestyle

behaviour contributing to diseases as involvement of females and men in the labour force are

significantly varied.

There are comparatively few literatures that studies the occupation and health

outcomes, especially health outcomes among different types of occupation. However, existing

literature is constrained to studying one or two health outcomes. Since the involvement of work

type involved by individuals due to occupational mobility obtained overtime to changes socio-

economic, technological in addition with attainment of education. It is now the right time to

study the effect of occupation on health, which will finally be our major concern as the amount

and diversities of disease reporting is increasing overtime. Through this paper I will extend my

study to more health outcomes over different occupation groups over different time

periods. Basically, I will be examining the structural change of occupation over time.

Diabetes and obesity are usually prevalent in developed countries. But recently, it is also

found that India has been slowly shifting towards the trend of having an obesity and diabetic

epidemic. These issues are of great concern in countries like the US and Canada. Thus, it is

important that India tries to control it before it turns out to be another impediment towards

development.

2. DATA AND SUMMARY STATISTICS

(^4) Hande Barlin and Murat Anil Mercan, “Occupation and obesity: effect of working hours on obesity by occupation groups,” Applied Economics and Finance 3, no. 2 (2016): 179-185. (^5) Margaret A. Allman-Farinelli, Tien Chey, Dafna Merom, and Adrian E. Bauman, “Occupational Risk of Overweight and Obesity: An Analysis of the Australian Health Survey,” Journal of Occupational Medicine and Toxicology 5, no. 1 (2010): 14, https://doi.org/10.1186/1745- 6673 - 5 - 14

3.1. DATA

In this study, I will utilize the National Family Health Survey (NFHS), which is a large-

scale, multi-round survey conducted in a large representative sample of households all over

India. This study employs two rounds of NFHS i.e., III & IV , where round III is surveyed

between 2005-2006 and round IV between 2015-2016. For the study I will be specifically

exploiting women recode and men recode merged with household recode. This paper basically

examines the effects of different groups of occupation on four major health outcomes such as

obesity, overweight, underweight and diabetes separately for women and men of the age group

range between 18 to 49 years over two time periods. The variables used for this study are as

follows:

Occupation Groups : Seven different categorical groups of occupation are being used to

analyse the effect on health outcomes. The groups are professionals or technical or managerial;

clerical; sales; agriculture; household and domestic services; skilled and unskilled manual

workers including a group of people not in the workforce or with no occupation. Agricultural,

household & domestic services and skilled & unskilled manual workers are grouped as blue-

collar jobs and professionals or technical or managerial, clerical, sales are being treated as

white-collar jobs by the convention of the paper. Also, these are the categories of major

independent variable for the estimation of this paper.

Body Mass Index (BMI) (weight in 𝑘𝑔/height×height in 𝑚^2 )^6 : It will be used as a measure of

determining three-main health outcomes such as obesity, underweight and overweight. Here in

this paper, obesity, underweight and overweight will be treated three different outcome

variables.

Obesity : According to WHO, people with BMI greater than 30 is considered as obese.^7 Obesity

will be taking value 1 if obese (i.e., BMI> 30) otherwise takes value 0 if not.

Overweight: According to WHO, people with BMI between 25-29.9 are considered as

overweight.^8 Overweight will be taking value 1 if overweight, otherwise takes value 0 if not.

(^6) Definition of BMI according to WHO: “Body Mass Index (BMI).” World Health Organization. Accessed November 16, 2023. (^7) Definition of Obesity according to WHO: “Body Mass Index (BMI).” World Health Organization. Accessed November 16, 2023. (^8) Definition of Overweight according to WHO: “Body Mass Index (BMI).” World Health Organization. Accessed November 16, 2023.

in agriculture, 3.5% household and domestic and 9.5% in manual works. In case of men ( Table

2 ) it is, 15.4% not in the workforce,7.3% in profession or technical or managerial, 4.4% in

clerical,12.8% in sales,23.3% in agriculture, 5.2% household and domestic, 31.6% in manual

work. When it comes to round 4, among women ( Table 3 ), 70.5% are not in the workforce,

2.8% professional/technical/managerial, 15% in agrarian, 6.3% manual work, 3.2% household

and domestic, 1.5% in sales and 0.4% in clerical. In men ( Table 4 ), it is 23.3% not in the

workforce,5.5% in profession or technical or managerial,1.8% in clerical, 8.6% in sales, 28.8%

in agriculture, 6.8% household and domestic, and 25.3% in manual work.

4. CROSS TABULATION

In round 3 of the women dataset ( Table 5.a ), it is seen that individuals in the

professional category are the most obese (having 5.43 % of the sample) and the least obese

individuals belong to the agriculture category (having 2.09%) as expected. The general trend

observed in this table is that occupations involving physical work have the least people obese

whereas the professions involving less physical labour have the most obese individuals. In the

case of underweight ( Table 5.b ), we see that 38.71% people involved in agriculture are

underweight, being the highest contributor to the sample and professional group being the

lower contributor to this section; having 17.05% only which is the expected result following

the previous deduction. However, what is interesting is that in the case of overweight ( Table

5.c), we see that women in the occupation of agriculture remain the highest contributor to this

category (95.99%) and the professional category being the lowest contributor with 76.29%.

This trend in 2005-06 highlights that as individuals involved in the white-collar jobs continue

to be more prone to be obese, the ones in blue collar jobs like agriculture also range from being

underweighted to overweight: safely avoiding the obese category. This is an interesting

observation and helps categorize professions to the range that they belong to.

In round 4, we see that the highest contributor to the obese category ( Table 7.a ) has

shifted from the professional category to the clerical category (8.43%). And the lowest

contributor to this category remains the occupation of agriculture. Hence it is clear that even

over time, we observe that occupations involving manual labour continue to positively

contribute to the individual’s health. However, the result obtained in the underweight section

( Table 7.b ) has not changed since the last round. The agriculture sector is still the most

contributor (26.99%) to the underweight category and the professional occupation is the lowest

contributor (11.31%) to this category. There continues to exist a stark contribution from the

agriculture sector (90.35%) to the overweight category ( Table 7.c ) and this result has reduced

slightly since the previous round. This is extremely interesting because first, the trend has not

changed and second, we would not expect them to be overweight given that they are engaged

in manual labour. However, what this means is that they belong to a range rather than a

particular section. It is also important to note that the method of agriculture depends on the

land holdings of the family which can give rise to different amounts of energy exerted, in turn

determining their BMI. Since the sample includes individuals from all sections, it might

indicate the large range of the sample. It could also be the case that since these percentages

are calculated relative to the number of observations, the difference in ‘n’ can lead to difference

in the percentage and thus the result.

Now if it comes to men, in round 3, we see that men belonging to the agriculture sector

are the least likely to be obese ( Table 6.a ) with only 0.64% being a part of this section and the

highest belongs to the professional sector (3.64%). This trend is similar to women from round

3. In the case of underweight ( Table 6.b ), we see that the highest percentage of the sample

belongs to the ones not in the workforce (43.18%) and the least percentage of the sample

belongs to the professional category as expected (12.38%). Here we see how men, who are not

in the workforce, are the most likely to be underweight. This might be because not being in the

workforce translates into not having sufficient income to eat healthy. And in the case of

overweight ( Table 6.c ), we observe that men belonging to the not in workforce, agriculture and

manual works category are the most overweight having about 90% as their contribution. The

trend that is recurring is that these three categories have observations that places them in the

range from overweight to underweight and avoid the obese category.

In round 4 of men, we see that those not in the workforce and agriculture (about 1.5%)

are the least obese category ( Table 8.a ) and professionals and clerical workers are the most

obese category (about 5.3%). In the underweight section ( Table 8.b ), we observe that men

belonging to not in the workforce form the most part of the sample (31.61%) whereas the least

part of the sample is formed by the professional category. This concept of the professional

sector not being underweight is directly linked with their type of occupation and the utilization

of their energy. Moving on to the overweight category ( Table 8.c ), we see that here too that the

not in workforce section forms about 90% of the sample followed by agriculture with 89%.

Surprisingly, the least percentage is observed in the professional category (72.68%). Thus the

“not in workforce” category men mainly belong to the range from underweight to overweight

and evidently avoid the obese category. Agriculture too falls under the same range. Relative to

round 3, we see that the range that these three categories, agriculture, not in the work force and

The estimation equation presented in Table 9 for women in round 3 gives the results as

follows: Women belonging to the agriculture sector are 1.48% points less likely to be obese

relative to the ones not in the workforce. Since 4.29% of the comparison group are obese

(comparable to the highest contributor to this category), it is a significant difference.

Additionally, as established in the cross tab, we find that they belong to the range of

underweight to overweight. That is, women in agriculture are more likely to be both

underweight and overweight by 2.69% points and 4.11% points respectively relative to the

ones who are not in the workforce. A similar trend is observed in the case of manual workers.

They are also less likely to be obese but are more likely to be both overweight and underweight.

What is interesting to observe is that after adding all the controls, we see that almost all

the occupations are negatively related to obese and range majorly between underweight and

overweight with some exceptions. Since an individual can belong to only one of these

categories, we see that the groups of occupations safely avoid the obesity category and settle

between being underweight and overweight. Even though this range is huge, it provides a

picture of the trend observed in the economy. Within the occupations, we observe that

occupations involving physical work are more likely to be underweight compared to the

occupations that have minimal movement. But no such trend is observed in the overweight

category.

To compare it with the results obtained in round 4 ( Table 11 ), we observe a similar

pattern where women belonging to most of the occupations categories are negatively correlated

with obesity. And they too range from being underweight to overweight. The trend in

agriculture has remained the same even after 10 years. But women belonging to the

professional category are more likely to belong to the overweight category (2.49% points

relative to not in the workforce) and less likely to belong to the underweight category (-0.40%

points). Similar to round 3, women involved in the blue-collar jobs are more likely to be

underweight relative to the ones in white collar jobs. The fact that this trend has not changed

in the last 10 years, given the increase in the usage of technology, expansion in the access to

processed food and the changes in lifestyle, it is clearly evident how significant the occupation

is on each individual’s BMI. Even with the increased awareness of healthy diet and the high

enrolment in gymnasiums, the occupation that one involves themselves in is still the major

determinant of the individual’s physical health.

This idea is further visible in the coefficient obtained in the urban variable as well. In

both the rounds, we see that if the individual belongs to the urban area, then they are more

likely to obese relative to the ones residing in rural areas (1.71% points in round 3 and 1.69%

points in round 4). Since the occupation type in the rural areas usually involves high manual

labour, there is a clear distinction in the probability of being obese.

Additionally, we see that the richest are the most likely to be obese by 3.25% points relative

to every other distinction in the wealth index in round 3. This relationship has significantly

increased in the last ten years to 6.54% points. This huge increase in probability can be

attributed to the increased consumption of processed food that includes trans fats which directly

contributes to obesity. Even though the rich have access to fitness machines, awareness about

the importance of a balanced diet and the access to healthy food items, it is not being translated

into a healthy lifestyle. This is one of the major drawbacks that has to be addressed.

In the religion variable in Round 3, even though they are not statistically significant,

we see that conditional on the individual belonging to the Christian or the Muslim religion,

there is an increased likelihood of obesity relative to the people belonging to the Hindu religion.

This tendency is observed in round 4 too. This pattern can be attributed to the meat-based diet

that is eaten by these groups. Since the meat-based diet has more fat molecules relative to the

plant-based diet which is usually followed by the people belonging to Hindu religion, we would

expect this relationship to be expressed.

The results for diabetes in round 3 and round 4 of women are as follows. In round 3,

women in the professional group are 0.63 % point less likely to be diabetics relative to not non-

working groups. In addition, women in agriculture and manual laborers are also 0.32 and 0.

percentage point less likely to diabetic. But when we look at round 4, we see that women in

clerical and agriculture are less likely to diabetic with 1.55 and 0.3 percentage points

respectively. Irrespective of the occupation type, women are less likely to be diabetic.

Furthermore, in both rounds, women in Muslim community are diabetics, mostly due to their

habit of eating. Also, women in urban communities are significantly diabetics in both rounds

with 0.38 percentage point relative to rural areas. Additionally, in round 4 it is seen that the

consumption of fried food has a positive likelihood of having diabetics as expected due to the

unhealthy diet of food.

Further, this paper could only account for the age group between 15-49 years old for

both men and women. It helped for the same level of comparison. But there might be people

above 49 years old who might be still in the workforce, who had to be included in the non-

workforce due to the framework of the equation variables. Moreover, diabetics are mostly seen

in individuals of age above 40. So, diabetics will be observed in people above 49 as well. This

results in reduction in data.

8. DISCUSSION AND CONCLUSION

We found that individuals involved in the white-collar jobs continue to be more prone

to be obese, the ones in blue collar jobs like agriculture also range from being underweighted

to overweight in all the rounds over time. Women falls under the category of underweight to

overweight safeguarding the from the pit of obesity. Men in both white-collar and blue-collar

jobs are less likely to be underweight and overweight relative to men not in the workforce with

significance. There is difference in obesity, underweight and overweight among men and

women. While observe that men are more diabetic than women, with an increase in the

percentage of diabetics’ overtime for both genders. Additionally, individuals in white collar

jobs seems are more diabetic than blue-collar jobs.

After accounting for all above results, it is relevant to study the impact of factors that

might have serious consequences on creating health outcomes other than occupation that an

individual is involved in. Because India is moving to an obesity pandemic. It will lead to other

consequential health outcomes. With the pace of growth of our economy, if we forget to

address these serious issues of obesity and other health outcomes, it will tumble down abruptly.

If the country faces health issues, the country will scare with a productive and healthy working

force which will later pull down the economy. Also, if it further exacerbates the situation, a

huge lump of money needs to be spent for the recovery of people. Thus, the government should

take actions before hands in terms of preventive measures. Some policies to ensure

adequate healthy food supply in the working space and facility for gyms or facilities for

physical building in the working space, especially in the white-collar job sectors.

Though people are educated at some level, people are not properly aware about the

consequences of not treating the body and health properly. How their ignorance in maintaining

their health acts back to their own impairment. So, the government should bring schemes and

awareness campaigns for the people to take preventive precautions to support the public.

On the whole, we see that the result found in this paper has a direct implication for the

country. After accounting for the said drawbacks, the found result can be utilized to devise

policy and avoid the worsening of the situation.

APPENDIX

• Table 1: Descriptive Statistics of Women in Round Section 1: Descriptive Statistics tables
  • Obesity 117060 .034 .182 Variable N Mean Std. Dev. Min Max
  • Underweight 117173 .29 .454
  • Overweight 117173 .868 .338
  • Diabetes 117173 .011 .104
  • Not in work force or no occupation 117173 .596 .491 Occupation Types
  • Professional or Technical or Man 117173 .04 .197
  • Clerical 117173 .013 .112
  • Sales 117173 .025 .155
  • Agriculture 117173 .196 .397
  • Household and domestic services 117173 .035 .184
  • Skilled and unskilled Manual work 117173 .095 .294
  • Urban 117173 .55 .497 Urban
  • Rural 117173 .45 .497
  • Age 117173 29.157 9.484
  • Poorest 117173 .113 .317 Wealth Index
  • Poorer 117173 .143 .35
  • Middle 117173 .193 .394
  • Richer 117173 .244 .43
  • Richest 117173 .307 .461
  • Scheduled Caste 115535 .833 .373 Caste
  • Scheduled Tribe 115535 .129 .335
  • OBC 115535 .038 .19
  • Hindu 116072 .733 .443 Religion
  • Muslim 116072 .133 .339
  • Christian 116072 .09 .286
  • Other 116072 .045 .208
  • No 117173 .985 .121 Smoking
  • Yes 117173 .015 .121
  • Drinks Alcohol 117173 .028 .165
  • Educations in Years 79857 3.949 1.636
  • State 117173 17.811 9.402
• Table
• Table 2: Descriptive Statistics of Men in Round
  • Obesity 63057 .016 .125 Variable N Mean Std. Dev. Min Max
  • Underweight 63097 .29 .454
  • Overweight 63097 .905 .293
  • Diabetes 63097 .012 .11
  • Not in work force or no occupation 63097 .154 .361 Occupation Types
  • Professional or Technical or Man 63097 .073 .26
  • Clerical 63097 .044 .205
  • Sales 63097 .128 .334
  • Agriculture 63097 .233 .423
  • Household and domestic services 63097 .052 .222
  • Skilled and unskilled Manual work 63097 .316 .465
  • Urban 63097 .496 .5 Urban
  • Rural 63097 .504 .5
  • Age 63097 29.622 9.71
  • Poorest 63097 .096 .295 Wealth Index
  • Poorer 63097 .14 .347
  • Middle 63097 .204 .403
  • Richer 63097 .262 .44
  • Richest 63097 .298 .457
  • Scheduled Caste 63097 .855 .352 Caste
  • Scheduled Tribe 63097 .118 .322
  • Other 63097 .027 .163
  • Hindu 62501 .746 .435 Religion
  • Muslim 62501 .125 .331
  • Christian 62501 .092 .29
  • Other 62501 .036 .187
  • No 63097 .68 .467 Smoking
  • Yes 63097 .32 .467
  • Drinks Alcohol 63097 .346 .476
  • Education in Years 63097 8.073 5.045
  • State 63097 19.192 9.439
• Table
  • Table 3: Descriptive Statistics of Women in Round
    • Obesity 671910 .042 .2 Variable N Mean Std. Dev. Min Max
    • Underweight 672734 .22 .414
    • Overweight 672734 .841 .365
    • Diabetes 672734 .014 .117
    • Not in work force or no occupation 116121 .705 .456 Occupation Types
    • Professional or Technical or Man 116121 .028 .165
    • Clerical 116121 .004 .062
    • Sales 116121 .015 .122
    • Agriculture 116121 .153 .36
    • Household and domestic services 116121 .032 .177
    • Skilled and unskilled Manual work 116121 .063 .242
    • Never 672734 .045 .208 Fried Food
    • Daily 672734 .115 .319
    • Weekly 672734 .338 .473
    • Occasionally 672734 .502 .5
    • Urban 672734 .71 .454 Urban
    • Rural 672734 .29 .454
    • Age 672734 29.8 9.752
    • Poorest 672734 .191 .393 Wealth Index
    • Poorer 672734 .213 .41
    • Middle 672734 .211 .408
    • Richer 672734 .198 .398
    • Richest 672734 .187 .39
    • Scheduled Caste 672734 .811 .392 Caste
    • Scheduled Tribe 672734 .143 .35
    • OBC 672734 .042 .201
    • Other 672734 .004 .061
    • Hindu 665303 .754 .431 Religion
    • Muslim 665303 .137 .344
    • Christian 665303 .071 .257
    • Other 665303 .037 .19
    • No 672734 .997 .057 Smoking
    • Yes 672734 .003 .057
    • Drinks Alcohol 672734 .024 .153
    • Education in Years 672734 6.719 5.188
    • District 672734 304.877 177.567
• Table
  • Table 4: Descriptive Statistics of Men in Round
    • Obesity 93398 .025 .155 Variable N Mean Std. Dev. Min Max
    • Underweight 93477 .2 .4
    • Overweight 93477 .859 .348
    • Diabetes 93477 .015 .123
    • Not in work force or no occupation 93477 .232 .422 Occupation Types
    • Professional or Technical or Man 93477 .055 .228
    • Clerical 93477 .018 .132
    • Sales 93477 .086 .28
    • Agriculture 93477 .288 .453
    • Household and domestic services 93477 .068 .252
    • Skilled and unskilled Manual work 93477 .253 .435
    • Never 93477 .074 .262 Fried Food
    • Daily 93477 .113 .317
    • Weekly 93477 .35 .477
    • Occasionally 93477 .463 .499
    • Urban 93477 .687 .464 Urban
    • Rural 93477 .313 .464
    • Age 93477 30.058 9.865
    • Poorest 93477 .17 .376 Wealth Index
    • Poorer 93477 .207 .405
    • Middle 93477 .216 .411
    • Richer 93477 .206 .405
    • Richest 93477 .201 .401
    • Scheduled Caste 93477 .19 .393 Caste
    • Scheduled Tribe 93477 .185 .388
    • OBC 93477 .414 .493
    • Other 93477 .211 .408
    • Hindu 92598 .785 .411 Religion
    • Muslim 92598 .108 .31
    • Christian 92598 .067 .25
    • Other 92598 .04 .196
    • No 93477 .853 .354 Smoking
    • Yes 93477 .147 .354
    • Drinks Alcohol 93477 .32 .467
    • Education in Years 93477 8.544 4.593
    • District 93477 313.523 182.844

CROSS TABULATION TABLES

Round 3: Women

Table 5.a

Table 5.a Cross tabulation of Occupation Types and Obesity of Women in Round 3 First row has frequencies and second row has row percentages

Table 5.b

Table 5b. Cross tabulation of Occupation Types and Underweight of Women in Round 3 Occupation Type of the individual Underweight Not Underweight Underweight Total Not in work force or no occupation 51021 18779 69800 73.10 26.90 100. Professional or Technical or Managerial 3922 806 4728 82.95 17.

Clerical 1246 255 1501 83.01 16.99 100. Sales 2372 526 2898 81.85 18.15 100. Agriculture 14095 8902 22997 61.29 38.71 100. Household and domestic services 3065 1037 4102 74.72 25.28 100. Skilled and unskilled Manual works 7456 3726 11182 66.68 33.32 100. Total 83177 34031 117208 70.97 29.03 100. First row has frequencies and second row has row percentages Occupation Type of the individual Obesity Non-obese Obese Total Not in work force or no occupation 66723 2990 69713 95.71 4.29 100. Professional or Technical or Managerial 4458 256 4714 94.57 5.43 100. Clerical 1422 79 1501 94.74 5.26 100. Sales 2739 157 2896 94.58 5.42 100. Agriculture 22868 128 22996 99.44 0.56 100. Household and domestic services 3936 165 4101 95.98 4.02 100. Skilled and unskilled Manual works 10940 234 11174 97.91 2.09 100. Total 113086 4009 117095 96.58 3.42 100.