The Crisis in India’s Agriculture Sector Which No One is Really Talking About

The slow growth of agriculture remains a weak link in the overall Indian economy. Let’s take a look at Figure 1. Figure 1 uses the new gross domestic product (GDP) series, which was launched in January 2015, to plot the GDP growth and the growth of agriculture, forestry and fishing, within that, over the last few years.

What does Figure 1 tell us? It tells us very clearly that the overall GDP growth is much faster than the growth in agriculture on the whole. There are exceptions to this rule. In the period of three months ending December 2013 and March 2014, growth in agriculture was a little faster than the overall GDP growth.

Figure 1: 

The trouble is that the new GDP series has data starting from only June 2011. Given this, we have growth rates from June 2012 onwards. In order to look at periods before June 2012, we need to use the old GDP series. On the Centre for Monitoring Indian Economy website, I could find data from December 1999 onwards.

I have plotted that data in Figure 2. Figure 2 plots the overall GDP growth along with the growth in agriculture, forestry and fishing.

Figure 2: 

Figure 2 like Figure 1 clearly tells us that even as per the old GDP series, the GDP growth as a whole was much faster than the growth in agriculture, forestry and fishing. Of course, there were occasions (though not many) where the overall GDP growth was slower than growth in agriculture.

The result of this slower growth has been the size of agriculture as a part of the overall economy has been shrinking over the years. This becomes clear from Figures 3 and 4. Figure 3 uses the new GDP series data, whereas Figure 4 uses, the old GDP series data.

Figure 3: 

Figure 4: 

What Figures 3 and 4 tell us very clearly is that the size of agriculture as a part of the overall economy (measured by the GDP) has shrunk rapidly over the years. This has happened because the overall GDP (i.e.the non-agriculture part of the economy) has grown at a faster rate.

As can be seen from Figure 4, in 1950-1951, agriculture formed more than 40 per cent of the economy. By now it is down to 12 per cent. This has happened as the average land holding on which farming is carried out has fallen over the years, as land has been divided across generations. The average size of the land farmed by the Indian farmer has fallen over the decades and in 2010-2011, the last time the agriculture census was carried out, stood at 1.16 hectares. In 1970-1971 it had stood at 2.82 hectares. This fall in farm size has made farming in many parts of the country, an unviable activity, leading to the size of agriculture as a part of the economy becoming smaller and smaller.

The trouble is that agriculture still employs a large portion of the workforce. While, it’s contribution to the overall economy has come down, a large number of people continue to remain dependent on income from agriculture. Take a look at Table 1.

Table 1:

YearAgriculture’s share in the rural economic output (in per cent)YearShare in rural employment (in per cent)

Source: Changing Structure of Rural Economy of India Implications for Employment and Growth, Ramesh Chand, S. K Srivastava and Jaspal Singh, Discussion Paper, NITI Aayog, December 2017.

As can be seen from Table 1, the share of agriculture in the rural economy has fallen from 72.4 per cent to 39.2 per cent, over the four-decade period between 1970-1971 and 2011-2012. During the same period, the proportion of people dependent on agriculture as a mode of employment, hasn’t come down at the same rate.

In 1972-1973, the share of agriculture in rural employment was 85.5 per cent. This came down to 64.1 per cent in 2011-2012. So, 64.1 per cent of the rural workforce produced 39.2 per cent of the rural economic output in 2011-2012. Hence, agriculture has many more people dependent on it, than is economically feasible.

It is clear from Table 1 that between 2004-2005 and 2011-2012, the proportion of people dependent on agriculture as employment has come down. This is a trend that would have continued. Even with that many more people continue to be dependent on agriculture than is economically feasible.

Also, given their lack of technical skills, it remains difficult for people trying to move out of agriculture to find other jobs. Or find jobs that pay better.

Table 2: Education level (general and technical) of usually employed rural workers of age 15-59 years

(per cent)
Percent of rural workersMaleFemalePersons
Secondary Education and above19.727.16.811.814.922.3
With technical education1.
With vocational training14.215.413.012.713.814.6

Source: Authors estimation based on unit-level NSS data on employment and unemployment survey.
Source: Changing Structure of Rural Economy of India Implications for Employment and Growth, Ramesh Chand, S. K Srivastava and Jaspal Singh, Discussion Paper, NITI Aayog, December 2017.

Table 2 tells us that the education and technical skills of rural India are fairly limited. Given this, their ability to find a job outside agriculture remains limited. Also, the secondary education which more than a quarter of men had had in 2011-2012, cannot be taken very seriously. Madhav Chavan, of the Pratham Education Foundation, estimates that in the period of the ten years up to 2015, 10 crore children completed primary school without the ability to do some basic reading and mathematics.

If people are to be moved out of agriculture, first and foremost the quality of education being imparted in rural India needs to improve. The focus, at least in the initial years, needs to be on teaching basic skills of reading, writing and the ability to do basic maths, rather than complete the syllabus, as is the case currently. Also, the current system of no exams till class VIII needs to change. Further, skill development in order to make individuals employable, needs to be encouraged further and happen at a much larger scale than it currently is.

There are a whole host of issues that are holding Indian education back. These need to be tackled with obvious as well as out of the box solutions. And for this to happen, the ministry of human resource development needs to be a high-profile ministry, in fact the most important ministry in the government, which it currently isn’t, nor was it earlier.

The column was originally published in Equitymaster on December 21, 2017.

Gujarat Elections: In 2018 and 2019, with More Socialism, Modi Will Become More Like Manmohan


As expected the Bhartiya Janata Party (BJP), won the state assembly elections in Gujarat. The margin of victory though left much to be desired.

Before the elections, the party president Amit Shah had talked about the party winning 150 out of the 182 seats in the state assembly. The party finally ended up with 99. So, there was a clear gap between expectation and reality. I say this because Amit Shah is a brilliant electoral strategist and his words should never be taken lightly.

Finally, it is the massive support that the BJP enjoys in the cities of Gujarat that pushed it through. In the four biggest cities of Gujarat, Ahmedabad, Surat, Vadodara and Rajkot, the party won 46 out of the 55 seats (Ahmedabad 16 out of 21 seats, Surat 15 out of 16 seats, Vadodara nine out of 10 seats and Rajkot six out of eight seats).

What this clearly tells us is that the Goods and Services Tax (GST) wasn’t as big an issue in the cities of Gujarat, as was made out to be in the days leading up to the elections, in the media. It clearly impacted a section of the population, but that wasn’t large enough to make an electoral difference. The only other explanation for this lies in the fact that even those impacted negatively by the GST, couldn’t get themselves to vote for the Congress.

The other interesting point here was that more than 5.5 lakh voters chose the NOTA (none of the above) option while voting. This amounted to 1.8 per cent of the total eligible votes. The NOTA votes were more than or close to the winning margins in nearly 24 constituencies. One explanation for this that has been offered is that a small section of the population which is unhappy with the BJP didn’t want to vote for Congress either.


The trouble with this explanation is that there is no way to verify it. It could even be the opposite.

Anyway, getting back to the point I was trying to make—the BJP won 46 out of the 55 seats in the four biggest cities of Gujarat. How did the electoral results look in the remaining 127 seats? The BJP won 53 of these seats. The Congress won 71. Hence, the Congress clearly did much better than the BJP beyond the four biggest cities.

There will be economic and political implications of these electoral results in other states like Rajasthan and Madhya Pradesh, where the elections are scheduled in the months to come. Some points are as follows:

a) The basic problems in India’s rural economy are not going to go away any time soon. The size of the average agricultural holdings in India has fallen as land has been divided across generations, making agriculture as a profession very non-remunerative.
Over and above this, India has too many people in agriculture than is economically feasible. A recent discussion paper by Niti Aayog points out that as of 2011-2012, agriculture employed 64 per cent of the rural workforce but produced only 39 per cent of the total rural economic output. Hence, for agriculture to be economically feasible 8.4 crore agricultural workers need to be shifted out of agriculture. This is around 70 per cent of the non-farm workforce in the rural areas. This isn’t going to happen overnight.

b) Of course, given this huge disguised unemployment in agriculture, people working in agriculture try to work in other areas as well. But the trouble is that there aren’t enough jobs going around for them. Data from the Labour Bureau suggests that only 52.7 per cent of the people looking for jobs all through the year, in rural India, are able to find one. Given this, nearly one in two people in rural India do not have jobs all through the year.

Hence, rural India has a problem at two levels: 1) Agriculture as a profession is no longer as remunerative as it used to be. 2) There are not enough other jobs, given their low skillsets, which people working in agriculture can take on, to add to their income over and above what they make in agriculture.

This explains why land-owning castes have been protesting all across the country. This includes the Patidar Patels of Gujarat.

c) Given this, the BJP in every state that it goes to election after Gujarat, it is likely to promise a farm loan waiver, as it has done in other states over the last one year. This is going to cost state governments all across the country a lot of money. It will also create moral hazard with future borrowers waiting for farm loan waivers than paying off their loans.

The question is why did the BJP not promise a farm loan waiver in Gujarat? The rural areas in Gujarat are not as badly placed as in other states. One reason for this lies in the fact that the livestock economy in the state, has continued to grow robustly. Also, over and above this, the non-farm economy in the rural areas, created job opportunities because of the overall faster growth in the state.

In fact, farm loan waive offs will become even more important given that, the states of Madhya Pradesh and Rajasthan, are not as urbanised as Gujarat is.

Also, in the run up to 2019 Lok Sabha elections, I see the minimum support price of agricultural crops going up. As per the Shanta Kumar Committee, the minimum support price system benefits under 6 per cent of the farming households in the country. While, increasing MSPs may not benefit many farmers, it does have a strong signalling effect.

d) The Modi government will also look to consolidate its position in the urban and semi-urban areas. And for that, chances are it will waive off Mudra loans of Rs 50,000 or lower. In total, over 7 crore of Mudra loans of less than Rs 50,000 have been given out.

e) As far as the Congress is concerned, it needs to start rebuilding itself, particularly in the rural areas because that is where its support is. This is a rather obvious insight.
To, conclude, the Modi government will give out doles and waive off loans, in order to improve its position. This strategy will not be much different from what the Congress led UPA government led by Manmohan Singh, did in the 2009 elections. This again goes with the broader point that I keep making—India has only one model of governance and that is the Congress model.

In the end, socialism will win. We will have a bigger government in areas that we really shouldn’t.

The column was originally published in Equitymaster on December 19, 2017.

India@70: Where are the jobs?

indian flag

On August 9, 2017, lakhs of people belonging to the Maratha caste poured into the city of Mumbai for a silent march. 57 similar marches had already taken place in the state of Maharashtra, starting from Aurangabad on August 9, 2016. This was the 58th. The rallying cause behind the marches was to protest against the rape and murder of a teenaged girl belonging to the caste in Ahmednagar district in July last year. Other than the rallying cause, the Marathas have demanded quotas in government run as well as aided educational institutions. They also want reservations in government jobs.

Marathas are not the only land-owning caste in the country demanding a reservation in government jobs. Similar demands have been made by the Patels in Gujarat, the Kapus in Andhra Pradesh, the Jats in Haryana and the Gujjars in Rajasthan. The question is why do land-owning castes suddenly want reservation in government jobs, seven decades after Independence?

A major reason for this lies in the fact that the average size of a farmer’s landholding has fallen over the years. As the State of Indian Agriculture Report of 2012-2013 points out: “As per [the] Agriculture Census [of] 2010-11, small and marginal holdings of less than 2 hectare[s] account for 85 per cent of the total operational holdings and 44 per cent of the total operated area. The average size[s] of [the] holdings for all operational classes (small & marginal, medium and large) have declined over the years, and for all classes put together it has come down to 1.16 hectare[s] in 2010-11 from 2.82 hectare[s] in 1970-71.”

Take a look at Figure 1.

Figure 1:  Decline in the average size of agricultural landholdings between 1970-1971 and 2010-2011.

Source: State of Indian Agriculture Report, 2012-2013.

The agriculture census is carried out every five years. Hence, the latest available data is as of 2010-2011. The situation would have only gotten worse since then. The trend of falling farm sizes can be clearly seen from Figure 1. As the same piece of land has got divided among more and more family members over the generations, the average holding has fallen dramatically. And this has made agriculture unviable for many in the land-owning castes. Hence, the demand for reservation in government jobs.

The trouble is that the government doesn’t create jobs anymore, neither at the level of state governments nor at the level of the central government. Hence, what will happen once the land-owning castes figure this out? Will they demand reservations in private jobs as well?

The rate of unemployment

The irony is that the huge demand for jobs among the land-owning castes and others is not reflected in India’s rate of unemployment. The Labour Bureau carries out the Annual Employment-Unemployment Survey. This Survey is hardly annual. It was first carried out in 2009-2010. It skipped a year and was carried out for the next three years. It skipped a year again in 2014-2015 and was carried out again in 2015-2016. The 2015-2016 Survey is what will be discussed here.

The Labour Bureau basically measures unemployment using two methods. The first method is called the Usual Principal Status (UPS) approach. In this approach, “the major time spent by a person (183 days or more) is used to determine whether the person is in the labour force or out of the labour force.”

As per this method, the rate of unemployment was just 5 per cent.

The second method is called the Usual Principal and Subsidiary Status (UPSS) approach. Here, “a person who has worked even for 30 days or more in any economic activity during the reference period of [the] past twelve months is considered as employed under this approach.” As per this method, only 3.7 per cent of the workforce was unemployed.

Such low rates of unemployment are hardly surprising given the definitions of unemployment that are being followed. In the first method, an individual might have been unemployed for close to half the year but would still be considered to be employed. In the second method, an individual might not have had a job for 11 months during the year and would be considered employed.

Given this, the rate of unemployment does not tell us anything about the desperate search for jobs. But there is another set of data points that the Labour Bureau puts out, and that rarely makes it to the media. Take a look at Table 1.

Table 1:  All-India percentage distribution of persons available for work for 12 months (UPSS approach).

Source: Report on the Fifth Annual Employment-Unemployment Survey, 2016.

Table 1 basically tells us what proportion of the population which is looking for a job all through the year is able to find one. Around 61 out of 100 Indians in the workforce looking for a job all through the year are able to find one. In rural areas, only around 53 out of 100 individuals who are looking for a job all through the year are able to find one. These numbers point towards the huge underemployment of India’s workforce.

This is hardly surprising given that in the last two financial years, agriculture has contributed around 14 per cent to the gross domestic product and employed close to half of the working population. There is a clear mismatch here. Around half the country’s workforce is only contributing 14 per cent of the GDP.

What this means is that there is huge disguised unemployment in the rural areas. Disguised unemployment essentially means that there are way too many people trying to make a living out of agriculture. On the face of it, they seem employed. Nevertheless, their employment is not wholly productive, given that agricultural production would not suffer even if some of these employed people stopped working.

So, the unemployment numbers might not point towards India’s distressing job situation but the underemployment number clearly does. This is also borne out in Figure 2, which has been sourced from a recent report titled OECD Economic Surveys India.

Figure 2:

This report puts the rate of unemployment among India’s youth between the ages of 15 and 29 at more than 30 per cent. These youths are neither employed nor in education or training.

Regular unemployment data

The Fifth Annual Employment-Unemployment Survey was carried out in 2015-2016. It has been close to a year and a half since then and we haven’t had any fresh unemployment data being published by the government.

As Volume 2 of the Economic Survey of 2016-2017 released earlier this month, points out: “The lack of reliable estimates on employment in recent years has impeded its measurement and thereby the Government faces challenges in adopting appropriate policy interventions.” It then lists out 10 ways used by the government to measure unemployment and the problems with them. The problems listed are: “Partial coverage, inadequate sample size, low frequency, long time lags, double counting, conceptual differences and definitional issues, rarely used for the purpose of employment estimation etc.” This, of course, leads to the question why have 10 wrong ways of measuring unemployment and not one right way?

The government has tried to correct this by setting up a task force headed by [now former] NITI Aayog Vice-Chairman Arvind Panagariya to generate timely and reliable employment data. This is a step in the right direction. The tragedy is that this should have happened many years back, even before Narendra Modi took over as the prime minister. Of course, the previous governments are to be blamed for this as well. The Modi government also took more than three years to initiate something to solve this problem.

The trouble is that close to one million Indians are entering the workforce every month. That makes it around 1.2 crore Indians a year. And the government is still struggling with counting the number of the unemployed.

What makes things worse is that most of the individuals who are entering the workforce are not skilled enough. Over the years, the government has tried to correct this by outsourcing skill development to the private sector rather than just depending on the Industrial Training Institutes or the ITIs. But the scale of operation continues to remain very small.

As the Economic Survey referred to earlier points out: “For urban poor, Deendayal Antyodaya Yojana National Urban Livelihoods Mission (DAYNULM) imparts skill training for self and wage-employment through setting up self-employment ventures by providing credit at subsidized rates of interest. The government has now expanded the scope of DAY-NULM from 790 cities to 4,041 statutory towns in the country. So far, 8,37,764 beneficiaries have been skill-trained [and] 4,27,470 persons have been given employment.” When one million Indians are entering the workforce every month, this is not even a drop in the ocean.

Other data points

While we may not know the right rate of unemployment on a regular basis, there is enough other data that suggests that job creation is not happening. Take a look at Figure 3. It basically plots the bank lending to industry.

Figure 3:

Source: Reserve Bank of India.  

The lending carried out by banks to the industry has fallen over the years. In fact, in 2016-2017, the lending to industry shrunk by more than Rs 50,000 crore. This basically means that on the whole, the banks did not lend a single new rupee to the industry in 2016-2017. The reason for this is very straightforward. The industry has defaulted on its past loans and banks are no longer in the mood to lend.

This also shows us that the industries are no longer borrowing and expanding and creating jobs in the process. Of course, banks are not the only source of borrowing for industry. If we were to look at the overall flow of financial resources to the commercial sector it was down by around 11 per cent in 2016-2017 in comparison to a year earlier (Source: RBI Monetary Policy Report April 2017).

Over and above this, demonetisation had a huge negative impact on jobs in the informal sector. The Bharatiya Mazdoor Sangh (a trade union affiliate of the BJP) estimated that nearly 2.5 lakh units in the unorganised sector were closed down. Then there is the latest Reserve Bank of India (RBI) Consumer Confidence Survey. More people now believe that the employment conditions have worsened over the last year.

The leaders of the Bharatiya Janata Party like to claim that crores of jobs have been created through Mudra (Micro Units Development and Refinance Agency Bank) loans given out by banks. In 2015-2016 and 2016-2017, a total of 7.46 crore individuals were given Mudra loans. Hence, 7.46 crore jobs were created is the logic that is offered. But this is something that the CEO of Mudra does not confirm. As he told NDTV recently, when asked how many jobs had these loans created: “We are yet to make an assessment on that… We don’t have a number right now, but I understand that NITI Aayog is making an effort to do that.

The point being India has a serious jobs problem and we aren’t doing much to tackle it. And there are going to be no acche din without jobs.

The column originally appeared on Newslaundry on August 15, 2017.

India’s Agriculture Crisis is Set to Become Worse


The gross domestic product(GDP) data for 2015-2016 was declared sometime back. As per this data, agriculture (actually agriculture, forestry and fishing), made up for around 14.1% of the GDP, during the course of the financial year. The trouble is that close to 50% of the population continues to depend on agriculture for a living.

This basically means that agriculture formed around one seventh of the Indian economy during the last financial year. At the same time around half of the population is dependent on it. The point being that it employs many people than it actually should. Hence, there is a huge disguised unemployment in the rural areas.

Disguised unemployment essentially means that there are way too many people trying to make a living out of agriculture. On the face of it they seem employed. Nevertheless, their employment is not wholly productive, given that agricultural production does not suffer, even if some of these employed people stop working

There are many more people than the sector requires and this leads to lower incomes for those who work in agriculture. The broader point is that if the average incomes need to go up, people need to be moved away from agriculture. But a new analysis suggests that this will not happen at the pace it was earlier expected to be.

Akhilesh Tilotia of Kotak Institutional Equities makes this point in a recent research note titled Forecasts of fewer jobs dull demographic sheen. Tilotia is also the author of The Making of India. He reviewed a “set of 24 industry reports commissioned by the National Skills Development Council (NSDC) and compare them with similar reports that NSDC had put together around the end of the last decade.”

The earlier reports had put the size of the Indian workforce at 65.4 crore by 2022. The number is now a lot lower at 57.5 crore. As far as number of people employed in agriculture in 2022 is concerned, the earlier estimates put the number at 11.4 crore or 18% of the workforce. As per new estimates the number of people who are expected to be working in agriculture in 2022, stands at 21.6 crore or around 38% of the workforce.

This basically means that nearly 10.2 crore more Indians will be dependent on agriculture as a mode of living, than was expected earlier. Further, by 2022, agriculture is expected to form around one-ninth of the GDP or the overall economic size of the country.

The automobile sector which was earlier expected to employ 4.8 crore individuals is now expected to employ only around 1.5 crore individuals. The same goes for the food processing sector, which was earlier expected to employ around 1.8 crore individuals, but is now expected to employ only 40 lakh individuals. On the other hand, the numbers for organised retail have gone up dramatically from 1.8 crore individuals earlier, to 5.6 crore individuals, as per the latest estimates.

Long story short, enough jobs will not be created to move people out of agriculture into other sectors where they can make a living.

In fact, as the Economic Survey of 2014-2015 points out: The data on longer-term employment trends are difficult to interpret because of the bewildering multiplicity of data sources, methodology and coverage. One tentative conclusion is that there has probably been a decline in long run employment growth in the 2000s relative to the 1990s and probably also a decline in the employment elasticity of growth: that is, a given amount of growth leads to fewer jobs created than in the past. Given the fact that labour force growth (roughly 2.2-2.3 percent) exceeds employment growth (roughly about 1½ percent), the challenge of creating opportunities will remain significant.”

As the Survey further points out:Regardless of which data source is used, it seems clear that employment growth is lagging behind growth in the labour force. For example, according to the Census, between 2001 and 2011, labour force growth was 2.23 percent (male and female combined). This is lower than most estimates of employment growth in this decade of closer to 1.4 percent. Creating more rapid employment opportunities is clearly a major policy challenge.”

One reason why enough jobs are not being created is because of what economists call falling labour intensity. Economic growth now generates fewer jobs in the non-farm sector (industry including manufacturing, construction, mining and utilities plus services sector) than it used to earlier. For every 1% increase in the gross domestic product, the non-agricultural employment went up by 0.52%, between 1999-2000 and 2004-2005. This fell to 0.38% between 2004-2005 and 2011-2012. (Source: D.Joshi and V.Mahambare, HIRE & LOWER–Slowdown compounds India’s job-creation challenge, Crisil Research, January 2014)

Hence, economic growth does not translate into the same number of jobs as it used to in the past. This basically means that economic growth is less labour intensive. This has happened primarily because of two reasons. First, the economic growth now is driven by less labour intensive sectors like business and financial services as well as information technology and information technology enabled services. These sectors require only one or two people to produce Rs 10 lakh of real value added Gross Domestic Product or economic output. This basically means that faster growth in these sectors does not necessarily translate into jobs. (Source: D.Joshi and V.Mahambare, HIRE & LOWER–Slowdown compounds India’s job-creation challenge, Crisil Research, January 2014).

This is clearly a big problem which does not have easy answers. Further, people dependent on agriculture are low on skill-sets that are needed for jobs in other sectors. It also needs to be pointed out here that moving people from agriculture into other areas is not so easy.

In fact, other countries which have grown at a very fast pace in the past, have experienced the same phenomenon. TN Ninan makes the point in The Turn of the Tortoise. Take the case of Thailand. Agriculture still constitutes close to 40% of its workforce. Or China, which has become the factory of the world. Around 35 per cent of the workforce is still engaged in agriculture, even though it produces just 10 per cent of the Chinese economic output.

The column originally appeared in the Vivek Kaul Diary on June 9, 2016