This article is part of The Wire‘s ‘India Black Boxed’ series. Read it here: Introduction | Part I
The current regime started with pretty impressive promises in 2014 with respect to employment (two core jobs a year). If realised, by March 2024, 40 crore new jobs should have materialised. Have they? If not, what is the reality?
The reversal of employment growth in non-farm sector and little progress in Skill India
The reality is the following. First, open unemployment was barely 2.1% in 2012 (the last year for which data was available before the BJP came to power). It had already nearly tripled to 6.1% in 2018 (National Survey Organization’s Periodic Labour Force Survey (PLFS), conducted annually since 2017-18), the highest rate in 45 years of India’s labour force surveys.
The total number of unemployed was one crore (2012) before the BJP came to power – but it had tripled by 2018 to three crore. The youth unemployment rates went through the roof for: those with middle school (class 8) education it rose from 4.5% to 13.7%; with secondary education (class 10) from 5.9% to 14.4%; those with higher secondary (class 12) education from 10.8% to 23.8%.
Educated unemployment worsened sharply. For graduates, the unemployment rate rose from 19.2% to 35.8%; and for post graduates from 21.3% 36.2%. All this did not deter the Government of India (GOI) from announcing a New Education Policy 2020 that higher education enrolment should rise from the prevailing 27% (for the relevant age cohort of 18-23 year olds) to nearly double to 50% by 2035. How are these new higher education graduates supposed to be employed, if the current crop of graduates face rising unemployment.
If anything, the India Skills Report 2021 argues that nearly half of India’s graduates are unemployable, i.e. education quality in our colleges/universities has deteriorated sharply, most noticeably after the massification of higher education in the last two decades. Two developments underlie this phenomenon. First, the number of affiliated colleges (attached to universities, where the exam is conducted by the state or central university) has grown in India from around 10 000 in the early 2000s to 42 000 to 2020. The Universities (let alone the University Grants Commission) have limited capacity to regulate or monitor the activities of such colleges; yet they have grown at a rate of 4 new colleges per day, without a weekend break. Two, most of these colleges are private, set up by builders and contractors, in connivance with local politicians, who are often elected to high offices. If this pattern of private college growth, the quality is unlikely to improve, so unemployment may remain much the same – if non-farm jobs do not grow fast enough to absorb new entrants.
Skill India has been flaunted as part of a new National Skills Policy in 2015. It laid out the goal of skilling 40 crore workers by 2022. The outcome? The workforce share that was formally vocationally educated/trained in 2012 was 2.3%. In 2022-23 that share had barely moved to 2.4%. Meanwhile, in administrative data reported by GOI for the Pradhan Mantri Kaushal Vikas Yojana (PMKVY): in its three versions, the total trained were 1.8 million, 4.5 million, and 0.4 million in PMKVY 1, 2, and 3. In each PMKVY, from 2018-2023, cumulatively 14%, 43% and 7% were placed in jobs. That tells us what the quantitative reach plus qualitative outcome is of these schemes. This is a short term training scheme (maximum 3-4 months), and a significant share of these “trained and certified” numbers are for Recognition of Prior Learning, lasting mostly two days at best.
Unemployment rates for those with formal vocational education had been 18.5% in 2012; it had shot up to 33% in 2018. For those with technical degrees unemployment rose from 18.8% to 37.3% in six years to 2018.
Yet, BJP can turn around and say that the share of youth voting for BJP in 2019 Lok Sabha elections was larger than in 2014 (CSDS Surveys). How do we explain this phenomenon? Two probable explanations apply: one, that the worsening of unemployment was too recent to not be internalised by youth by 2019; two, the GOI adroitly did not allow the 2017-28 PLFS data of unemployment to be released before 2019 elections results were out, but released the data a week after it had won national elections. (Of course, Pulwama-Balakot nationalism likely trumped any discomfort unemployment was causing.)
This decade current jobs crisis is deepening for three large groups…
There are three groups who need non-farm jobs, whose numbers are constantly growing – so we have both a high stock plus rising flow problem with unemployment. First, the stock of unemployed in 2023 October is 42 million (at 10% of the workforce, according CMIE), a stock to which every month a flow of a few hundred thousand get added. Second, there is a regular flow of young people (turning 15 or over) who are looking for work, who number at least eight million (or 80 lakhs), and their numbers grow each year, as the share of the working age population rises at an accelerating pace each year (at least till 2030, after which it will still rise but at a decelerating pace). An important and growing share of these young are girls, whose educational levels have grown sharply in the last two decades, as demand for education rose. India achieved secondary enrolment of 85% for 15-16 year olds in 2015 (rising from 58% in 2010), with gender parity.
A third group in need of jobs is a stock of under-employed workers in agriculture, constituting 46% of the WF in 2023, who contribute only 15% of India’s GDP. In 1970 the average farm plot size was 2.25 hectares; which fell, due to rising population to 1.25 ha in 2010, and continues to fall. These plot sizes are too small to achieve economies of scale, and adopt new technical inputs, so productivity and yields on these farms remain low, and hence drives down agri-incomes, forcing migration for non-farm livelihood.
Tragically, government policies since 2016 have not only reduced the number of new non-farm job creation till Covid. More importantly, starting April 2020, after a sudden, national lockdown of very high stringency (by international metrics, as determined by Oxford University’s Blavatnik School of Governance), millions were forced to migrate back to migrant ‘origin’ states from ‘destination’ states: a dramatic six crore (60 million) workers were added to agriculture between April 2020 and June 2023.
…compounded by reversal of structural change: Agriculture rising and manufacturing falling
Structural change in India should consist of not only the share of industry in GDP rising, but also the share of workers in agriculture falling. In India we have seen the opposite since 2020: the share of agriculture has risen, and even worse, the absolute number of farm workers, which began falling for the first time since independence from 2004 (on account of rapid growth and rapid non-farm jobs growth), and which continued till 2019, has been reversed in a matter of <3 years.
This constitutes a reversal of the process of structural change, but it is the opposite of what was promised by GOI on assuming power: increasing the share of employment in manufacturing by 100 mn and a rise in share of its contribution to GDP from 17% to 25%. Two points are in order here. The first is that for 25 years preceding 2015, the share of manufacturing in GDP had remained stable at around 17%; thanks to poorly designed and badly implemented policy decisions (demonetisation at four hours notice of 86% of Indian currency, a national Goods & Services Tax with five rates introduced without adequate systems planning) manufacturing began shrinking from 2016 onwards, falling to all time low of 13% of GDP in the next four years. Simultaneously, labour intensive manufacturing particularly collapsed, mainly in the unorganised sector. This happened despite the hype about ‘Make in India’.
Employment in manufacturing also collapsed: it fell not only as a share of total employment (from 12.8% in 2012 to 11.5% in 2018), which is way below that of Bangladesh (16% of employment). This completed the reversal of structural change which will take years to correct. While manufacturing employment has only risen by 2022 to just above the level of total manufacturing employment prevailing in 2012, manufacturing contribution to GDP has also only just climbed back up to pre-Covid levels (17%) – a lost decade for structural change.
The second point is that this reversal was happening with economic growth slowing to 5.7% over the last 9 years compared to 8% pa over 2014-14. This slowdown is structural, mostly already was occurring pre-Covid. There are several signs of this slowdown. First, the number of job growth in non-farm employment had dropped to 2.9 mn (or 29 Laks) per year over 2013 to 2019 from 7.5 million (75 lakhs) pa between 2004-5 to 2012. Due to fast non-farm job growth in the earlier period, real wages were rising, while they have been flat both between 2013-2019 as well as since Covid till 2023. Hence, private consumption – the first driver of growth – has only been maintained by cuts in household savings (which have fallen as a share of GDP from 24% to 20%), and retail-level borrowing from banks has increased household debt levels. There are limits to sustaining private consumption based on expenditures by middle/upper classes.
The second driver of growth – investment to GDP – which was 31% (2013-14) when the current government came to power – has never recovered to that level in the last decade. It fell to 26% and is only just reached 29% in 2022-23. Public investment – has been unable to compensate for stagnant private investment (relative to GDP).
The third driver of growth – exports – actually fell for the first time after 2014, for merchandise goods from its $318 billion level in 2013-14 – for the next five years, not recovering that level till the sixth year after 2014. It was held up by a rise in services exports.
The fourth driver of growth – government expenditure – had been in consolidation mode to bring the fiscal deficit down, as the slowing GDP growth was in turn slowing revenue growth. Hence, the GOI faced a silent fiscal crisis prior to Covid. Yet GOI had no hesitation in reducing in one go the Corporate Tax Rate from 30 to 25% in 2019, losing 1.45 lakh crore per annum in CIT revenues. Not surprising that debt to GDP has risen from Rs 55 lakh crore in 2014 to Rs 172 lakh crore now; shooting up from <60% to now 81% of GDP.
These dimensions of the structural crisis are keeping growth, after the rebound from the contraction of 6.6% in FY21, to just around 6% going forward (according to the IMF). This is nowhere close to the minimum of 8% pa needed to reduce poverty, let alone the 10-12 million new non-farm jobs needed every year.
Not surprising, the GOI economists keep spinning new narratives about job growth in the last 10 years, every few months. The latest narrative is from HSBC, Morgan Stanley and S&P economists/equity analysts about the “New India” economy, based on a) digital infrastructure, based on the laudable India Stack; b) fin-tech, e-commerce, ed-tech, logistics, and capital attracted to promote start ups. However, this might at best be 15% of the economy. How this 15% will pull up the remaining 85% to take India to 8% pa GDP growth is left to the imagination.
The second effort of spin doctors is to trash CMIE data on employment – which is showing a 10% unemployment rate in October 2023, when the PLFS 2022-23 says it is 5.3% (according to Current Weekly Status, and lower by UPSS). The trashing of CMIE happens despite the following: the CMIE data uses internationally compliant definitions (which PLFS does not); its sample is actually larger than that of NSO’s PLFS; and it covers both the organised an unorganised sectors, and both rural and urban areas.
The lack of ILO-compliant definitions by NSO has the following effect: NSO regards unpaid family labour as employment, which CMIE does not. Nearly 100 countries of the world do not treat unpaid family labour as employment, and despite the Standing Committee on Economic Statistics (I was a member 2020-2022), an expert body advising NSO for the last several years to revise its stand, it failed to do so. The result: the >50 million increase in unpaid family labour in the last 6 years (as NSO shows) leads to a rising Labour Force Participation Rate and Worker Participation Rate, along with falling unemployment – the opposite of what the CMIE data shows, correctly. Hence, government economists continue to spin the narrative that employment has recovered post Covid.
A third effort of government economists is to spin a narrative is that EPFO new registrations since 2017-18 are growing, so organized jobs have grown. No serious economist outside the government considers EPFO registrations as an indication of new jobs, as opposed to, in some cases of some formalization of existing jobs.
A fourth effort such economists is to claim that any comparison of PLFS data (which started from 2017-18) with earlier NSO labour force surveys is not valid, as they are incomparable. We have shown above using the same comparison, that employment and wages were growing rapidly during 2004-12, compared to the period since then. However, this is about the silliest of the efforts to sell the government narrative, since even the NSO has in its 2017-18 compared its data to the period going back to 2004-5.
A fifth narrative is that MUDRA and SWanidhi loans have created millions of informal jobs. It is true that self-employment share of total employment has grown from 52% to 58% in five years; but as we have seen, most of this increase is unpaid family labour, working with some own-account worker (whose numbers have also grown phenomenally). All this increase in self-employment is at the regular employment, which is most secure form of work. MUDRA loans, 95% of which are Shishu loans of <Rs 50k, of which the average loan size is Rs 29K – a loan size barely enough to survive. Naturally, their default rate is high. It seem to have been the GOI’s way to get commercial banks to compensate people for losing jobs, and make ‘pakoras’, while involving their spouses in unpaid family labour.
All in all, GOI efforts to spin a jobs growth narrative has meant the following: it is not willing to recognise a glaring problem, and hence no concrete efforts are needed to change economic policy to make the growth pattern more labour-intensive. This means we are confronted with ‘lost decade’ of jobs, with barely 15-17 years left for the demographic dividend to end: a once in a lifetime opportunity in the life of any nation
Santosh Mehrotra is Research Fellow, IZA Institute of Labour Economics, Bonn.