COVID-19 induced recession worst since the Great Depression

In the latest World Economic Outlook (April 2020), the IMF projected global growth to contract sharply to -3% in 2020. The cumulative loss to global GDP over 2020 and 2021 from the pandemic crisis could be around 9 trillion dollars, greater than the economies of Japan and Germany, combined. For the first time since the Great Depression both advanced economies and emerging market and developing economies are in recession. It argues that many countries face a multi-layered crisis comprising a health shock, domestic economic disruptions, plummeting external demand, capital flow reversals, and a collapse in commodity prices.

Under the assumption that the pandemic and required containment peaks in the second quarter for most countries in the world, and recedes in the second half of this year, in the April World Economic Outlook the IMF projects global growth in 2020 to fall to -3 percent. This is a downgrade of 6.3 percentage points from January 2020. This makes the Great Lockdown the worst recession since the Great Depression, and far worse than the Global Financial Crisis.

Assuming the pandemic fades in the second half of 2020 and that policy actions taken around the world are effective in preventing widespread firm bankruptcies, extended job losses, and system-wide financial strains, the IMF projects global growth in 2021 to rebound to 5.8 percent.

In an adverse scenario, the pandemic may not recede in the second half of this year, leading to longer durations of containment, worsening financial conditions, and further breakdowns of global supply chains. This could mean even greater fall in global GDP: an additional 3 percent in 2020 (below the baseline in 2002) if the pandemic is more protracted this year, while, if the pandemic continues into 2021, it may fall next year by an additional 8 percent compared to our baseline scenario.

The IMF recommends that countries continue to spend generously on their health systems, perform widespread testing, and refrain from trade restrictions on medical supplies. It also recommends continued support to households and businesses throughout the containment period: credit guarantees, liquidity facilities, loan forbearance, expanded unemployment insurance, enhanced benefits, and tax relief. During the recovery phase, policies should shift to supporting demand, incentivizing firm hiring, and repairing balance sheets in private and public sector. It recommends moratoria on debt repayments and debt restructuring to be continued during the recovery phase too. 

Substantial targeted fiscal, monetary, and financial measures to maintain the economic ties between workers and firms and lenders and borrowers is required to keep intact the economic and financial infrastructure of society. Meanwhile, broad-based stimulus and liquidity facilities to reduce systemic stress in the financial system can lift confidence and prevent an even deeper contraction in demand by limiting the amplification of the shock through the financial system and bolstering expectations for the eventual economic recovery.

INDIA: 

• The IMF has projected India’s economy to grow at 4.2% in 2019/20, 1.9% in 2020/21 and then quickly recover to 7.4% in 2021/22. 
• Inflation is projected to remain low at 3.3% in 2020/21 and 3.6% in 2021/22. 
• Current account balance is projected to narrow to -0.6% of GDP in 2020/21 and then increase to -1.4% of GDP in 2021/22.

NEPAL:

• The IMF has projected Nepal's economy to grow at 2.5% in 2019/20 and 5% in 2020/21. 
• Inflation is projected to remain at 6.7% in FY2020 and FY2021. 
• Current account balance is projected to be -6.5% in FY2020 and -6.2% in FY2021.

Policy measures for the immediate-term:

• Sizable, specific, temporary, and targeted fiscal measures to cushion the impact on the most affected households and businesses, and to preserve economic relationships (by reducing firm closures). Digital payments may improve the delivery of targeted transfers to the informally employed.
• Central banks can provide ample liquidity to banks and nonbank finance companies, and credit guarantee on loans to SMEs. It could also encourage banks to renegotiate loan terms for distressed borrowers without lowering loan classification and provisioning standards. Beyond conventional interest rate cuts, expanded asset purchase programs may be helpful. 
• Broad-based fiscal stimulus such as public infrastructure investment or across-the-board tax cuts can help to boost confidence, stimulate aggregate demand, reduce bankruptcies and avert an even-deeper downturn. 
• Flexible exchange rates should be allowed to adjust as needed. Temporary capital flow measures on outflows could be useful.

Policy measures for the recovery phase:

• Secure swift recovery as even after the containment phase, uncertainty about contagion could subdue consumer demand. Firms will hire staff and utilize capacity gradually. Hiring subsidies may be required and worker retraining programs may alleviate labor market friction. Scaling back targeted, temporary measures may be warranted. 
• Balance sheet repair and debt restructuring may be required. Banks and regulators should encourage early and proactive recognition of nonperforming loans. Steps to strengthen the insolvency and debt enforcement framework, and measures to facilitate the development of a distressed debt market could be helpful.
• Strong multilateral cooperation is a must, especially on international trade and multilateral assistance.

GDP growth to decline 2.8% in FY2020 in Nepal and in FY2021 in India

In its latest South Asia Economic Focus (April 2020), the World Bank projects a sharp economic slump in South Asia, caused by halting economic activity, collapsing trade, and greater stress in the financial and banking sectors. The WB estimates that regional growth will fall to a range between 1.8 and 2.8 percent in 2020, down from 6.3 percent projected six months ago. Prolonged and broad national lockdowns will push growth in the negative territory. In 2021, growth is projected to hover between 3.1 and 4.0 percent, down from the previous 6.7 percent estimate. 

The WB notes that “the impact of the pandemic will hit hard low-income people, especially informal workers in the hospitality, retail trade, and transport sectors who have limited or no access to healthcare or social safety nets.” It recommends stablishing temporary work programs for unemployed migrant workers, enacting debt relief measures for businesses and individuals, and easing inter-regional customs clearance to speed up import and export of essential goods during the short-term. Over the medium-term, it recommends expansionary fiscal policies combined with monetary stimulus to keep credit flowing in their economies. 

Because of the drying up of tourism, disruption of supply chains, collapsing demand for garments, deteriorating consumer and investor sentiment, withdrawal of international capital, and decreasing remittance inflows, the effect of COVID-19 on South Asian economies is particularly severe. The report argues that the crisis has increased inequality in South Asia because the poor people have a higher likelihood of losing work, and domestic migrant workers are being pushed back to rural areas from where they moved to cities to escape poverty. Food security risks have increased. 

Here is an update on GDP forecast for FY2020:

For Afghanistan a deep recession is expected this year, with a contraction between 3.8 and 5.9 percent. With population growth of 2.3 percent, this implies a dramatic drop in per capita income.

In Bangladesh, with a population growth of 1 percent per year, a limited increase in per-capita GDP is projected for two years. That would be an abrupt change from high growth rates in recent years. Given the variation within the country, it means that significant parts of the population would lose income during these two years. GDP is expected to growth at 2.0% to 3.0%.

In Bhutan, growth is still expected, but the downward revision from Fall forecast is large in both years.

In India, GDP growth in FY2021 is expected to range between 1.5 and 2.8 percent, implying per-capita GDP growth of between 0.5 and 1.8 percent. In FY2020, GDP is expected to grow at 4.8% to 5%.

Most affected is the Maldives, where tourism directly and indirectly contributes two-thirds of GDP, 80 percent of exports and 40 percent of revenues. A contraction of the economy between 8.5 and 13.0 percent is expected in 2020. With population growth of 1.8 percent in 2019, the per-capita income loss will be significant.

Nepal, with population growth of 1.1 percent per year, would experience low per-capita growth for two years in a row. GDP is expected to grow at 1.5% to 2.8% in FY2020. In FY2021, GDP is expected to grow at 2.7% to 3.6%.

For Sri Lanka a recession is anticipated, with annual growth estimated between -3.0 and -0.5 percent.

Pakistan, which has already experienced low growth rates in recent years, could well fall into a recession. With 1.8 percent population growth, that would imply a painful decline in per-capita income.

Economic impact of COVID-19, policy measures required and how to finance deficit in India

Adapted from McKinsey & Company's latest brief on Getting ahead of coronavirus: Saving lives and livelihoods in India

In scenario 1, the economy could contract by about 10 percent in the first quarter of fiscal year 2021, with GDP growth of 1 to 2 percent in fiscal year 2021. In this scenario, the lockdown would be relaxed after April 15, 2020 (when the 21-day deadline is due to expire), with appropriate protocols put in place for the movement of goods and people after that. Our economic modeling suggests that even in this scenario of relatively quick rebound, the livelihoods of eight million workers, including many who are in the informal workforce, could be affected. In other words, eight million people could have their ability to subsist and afford basic necessities, such as food, housing, and clothing, put at severe risk. And with corporate and micro-, small-, and medium-size-enterprise (MSME) failure, nonperforming loans (NPLs) in the financial system could rise by three to four percentage points of loans. The amount of government spending required to protect and revive households, companies, and lenders could therefore be in the region of 6 lakh crore Indian rupees (around $79 billion), or 3 percent of GDP.

In scenario 2, the economy could contract sharply by around 20 percent in the first quarter of fiscal year 2021, with –2 to –3 percent growth for fiscal year 2021. Here, the lockdown would continue in roughly its current form until mid-May 2020, followed by a very gradual restarting of supply chains. This could put 32 million livelihoods at risk and swell NPLs by seven percentage points. The cost of stabilizing and protecting households, companies, and lenders could exceed 10 lakh crore Indian rupees (exceeding $130 billion), or more than 5 percent of GDP.

Scenario 3 could mean an even deeper economic contraction of around 8 to 10 percent for fiscal year 2021. This could occur if the virus flares up a few times over the rest of the year, necessitating more lockdowns, causing even greater reluctance among migrants to resume work, and ensuring a much slower rate of recovery.

What policy measures are required?

[...]Several measures have already been announced to provide liquidity, limit the immediate NPL impact, and ease personal distress for needy households in India. These amount to around 0.8 percent of GDP. Additional measures could be considered to the tune of 10 lakh crore Indian rupees, or more than 5 percent of GDP in fiscal year 2021. All the estimated requirements may not necessarily be reflected in the fiscal deficit of the current year—for example, some support may be structured as contingent liabilities that only get reflected when they devolve. However, a package of this order of magnitude may be essential in supporting those dealing with the possible steep declines in aggregate demand and in protecting the financial system from the possible solvency and liquidity risks arising from stressed companies if scenario 2 or scenario 3 plays out.

[...]Consideration could be given to an income-support program in which the government both pays for a share of the payroll for the 60 million informal contractual and permanent workers linked to companies and provides direct income support for the 135 million informal workers who are not on any form of company payroll. India’s foundational digital-identity infrastructure, Aadhaar, enables effective mechanisms for direct support, including through the Pradhan Mantri Jan-Dhan Yojana (PMJDY) and Pradhan Mantri Kisan Samman Nidhi (PM-KISAN) programs and to landless Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) beneficiaries. Concessions for home buyers, such as tax rebates for a time-bound period, could stimulate the housing market and unlock the job multiplier.

For bankruptcy protection and liquidity support, MSMEs could receive liquidity lines from their banks, refinanced by the Reserve Bank of India and a loan program for first-time borrowers could be administered through SIDBI.3 Substantial credit backstops from the government could be instituted for likely new NPLs Timely payments to MSMEs by large companies and governments could be encouraged by promoting bill discounting on existing platforms.

For large corporations, banks could be allowed to restructure the debt on their balance sheets, and procedural requirements for raising capital could be made less onerous. The Indian government could consider infusing capital through a temporary Troubled Asset Relief (TARP)-type program (such as through preferred equity) in a few distressed sectors (such as travel, logistics, auto, textiles, construction, and power), with appropriate conditions to safeguard workers and MSMEs in their value chains. Banks and nonbanks may also require similar measures to help strengthen their capital, along with measures to step up their liquidity and the liquidity in corporate-bond and government-securities markets.

How to finance?

Given that India’s fiscal resources are constrained, the Reserve Bank of India may need to finance a portion of such incremental government spending. The spending could be tracked as a COVID-19 portion of the budget to boost transparency. The inflationary effects may be low, as lockdowns severely constrict demand and the fiscal support provided would be a substitute for expenditure rather than additional stimulus. Price increases could, however, occur in some sectors, such as food, so appropriate steps would be needed to maintain harvests and keep the food supply chain operating smoothly.

Overall, devising a credible, systemwide, stabilization package would benefit from being executed in a timely fashion so it can influence the pace of recovery and help avoid severe damage to livelihoods, the economy, the financial sector, and society.

Following the first wave of stabilization measures, attention could shift to implementing the structural reforms needed to increase investment and productivity, create jobs quickly, and improve fiscal health. This could mean introducing further reforms in infrastructure and construction and accelerating investments in health, affordable housing, and other urban infrastructure. States could accelerate spending, and institutions such as NIIF4 could deploy domestic and long-term foreign capital faster. Such reforms could also enable Make in India sectors to become globally competitive and boost exports (such as electronics, textiles, electric vehicles, and food processing), strengthen the financial sector, deepen household financial savings and capital markets, and accelerate asset monetization and privatization to raise resources.

Stimulating demand in India through construction sector

Raising demand through construction sector stimulus

Ajay Shankar writes in Financial Express: [...]The construction sector has a large multiplier effect and is labour-intensive. So, it is a natural choice. There are a large number of incomplete housing projects with developers in difficulty/bankruptcy. The FM had announced a financing package last year to mitigate the economic downturn. But progress has been slow. There were issues with the fine print of the sanction orders. A simple but bold approach could work immediately; a takeover of all the incomplete projects from the developers, and getting the banks to immediately provide financing for completion at current costs with a government guarantee, could work. A czar could be designated with a mandate to get actual work started within 45 days of the end of the lockdown, and completion within 18 months of the commencement of work.

Preparatory work for takeover, tying up finances and settling contractual terms with the construction agencies can be done now. This would not need budgetary outflows. As the economy recovers and demand for housing picks up later, then the land bank with the developers would become liquid assets and debt may be comfortably serviceable.

A step up in the ongoing affordable housing construction programme could also generate additional demand. The same would also apply to the rural road programme. The present crisis has highlighted the inadequacy of hospital care capacity. While efforts are on to create temporary additional capacity on a makeshift basis, there is a need to increase capacity by 25-50% on an urgent basis. This merits funding from the stimulus package. This would generate demand for the construction industry as well as for the supply of medical equipment and furnishings.

Food subsidy, fiscal conundrum and mapping of migrant workers in India

Excess stocks of the Food Corporation of India must be released to the poor

Jean Dreze writes in The Indian Express: Everyone knows that the country has large food stocks, and that some of this could be used to protect people from hunger during the coronavirus crisis. The enormity of the situation, however, has escaped many observers. [...] Last year, in June (when the stocks normally peak), foodgrain stocks crossed 80 million tonnes — more than three times the buffer-stock norms. This year, they have reached a staggering 77 million tonnes in March, before the rabi harvest, when food stocks typically rise by another 20 million tonnes or so. Public food storage on this scale has never happened in India before. Meanwhile, the shadow of hunger looms large as the lockdown devastates people’s livelihoods. The finance minister did not do them a big favour by doubling PDS rations for the next three months — something like that was needed, in any case, to make space in FCI’s overflowing godowns for rabi procurement. A serious relief package would include releasing excess stocks to the states in large quantities.

Why is it proving so difficult? One reason has to do with food-subsidy accounting. The food subsidy essentially pays for the losses FCI makes when it buys at minimum support prices and sells at much lower PDS prices, and also the money spent on transportation and storage. As it happens, however, the food subsidy does not enter the central government’s accounts until stocks are released. That is why the finance minister had to budget Rs 40,000 crore in her relief package simply to release some excess food stocks into the PDS. In economic terms, releasing excess stocks is costless, and even saves money. But in accounting terms, it is expensive. This anomaly makes it harder to release food stocks: Credit-rating agencies watch the fiscal deficit, not the food economy.

[...]In short, food transfers are bound to play a big role in keeping poor people alive in the next few months. Food schemes such as the PDS and mid-day meals are in place in most villages, it is mainly a matter of reinforcing them. For this to happen, the central government must unlock the godowns and give plenty of food to the states. Never mind if the step takes the fiscal deficit a notch higher due to muddled accounting.

Devesh Kapur and Arvind Subramanian outlined five measures to secure fiscal space needed to address the economic crisis wrought by the COVID-19 pandemic.

---

Govt begins mapping of migrant workers for relief measures

From Business Standard: The central government has begun one of the most comprehensive exercises to map migrant workers scattered across the country — in relief camps, on their employers’ premises, or in clusters where they reside. The government wants to create a database of millions of such workers to ascertain whether a relief package could be announced for the most affected segment of the workforce due to the national lockdown to contain the spread of coronavirus (Covid-19), a senior labour and employment ministry official said. The Union home ministry and the labour ministry have asked state governments to coordinate with the chief labour commissioner’s (CLC’s) office to give a comprehensive data of all the migrant workers by April 11.

[...]According to the central government’s estimates, part of its response to a petition filed by activists Anjali Bhardwaj and Harsh Mander in the Supreme Court, around 1.03 million people are residing in relief camps. But this might be an underestimation because the information was not captured from all the shelter homes. Additionally, at least 1.5 million workers are being provided shelter by employers across the country.

[...]According to official estimates, 500,000-600,000 workers had to walk back home on foot because public transport was not available to them. They travelled miles on foot to reach their villages. Hundreds of thousands of migrant workers are still living in shelter homes set up by various state governments in India, while the rest are under quarantine facility before they are allowed to meet their families.

Migrant crisis in India: 

• 0.5 to 0.6 million workers walked on foot to villages after lockdown
• 8.4 million were given food by government and NGOs
• 1.03 million are in relief camps or shelter homes
• 1.5 million given shelter or food by employers
• 22,567 shelter homes (Kerala accounts for 70% of them)

Financing COVID-19 related deficit in India and the impact of COVID-19 on Nepali economy

Fiscal space: Not if but how

In an op-ed published in Business Standard today (ungated version here), Devesh Kapur and Arvind Subramanian argue that the Indian government will have to find the funds/revenue to respond to the economic crisis wrought by the spread of COVID-19 pandemic. The question of fiscal space in India is not about 'if' but 'how'. They propose five ways of financing additional expenditures over the next 12 months:

• Reduction in other expenditures (Rs 1-1.5 trillion): Cut recently initiated projects and fund those near completion. Don't spend money in reviving poorly functioning public enterprises.
• Foreign borrowing, from official sources and non-resident Indians (NRIs; Rs 1-1.5 trillion): Borrow from multilateral banks (WB, ADB, etc) and also ask them to repurpose existing loans. Make a contingency plan for seeking quick disbursing funds from the IMF. Tap NRIs on special bonds.
• Public financing by issuing g-secs (including to banks and LIC) (Rs 5 trillion): Raise moeny from the public by conventional bond issuance. Allocate some of the new borrowing to PSBs and LIC.
• Monetary financing or “printing money” (Rs 1-1.5 trillion): Ask RBI to directly buy government securities and state government bonds. Make this direct purchase a one-off event given the exogenous shock (i.e, shock is not due to fiscal profligacy). This should not shoot up inflationary pressures and would also not incentivize government to monetize fiscal deficit repeatedly.
• Mobilizing additional resources via raising taxes and cutting subsidies (Rs 1-1.5 trillion)

---

COVID-19 sledgehammers economic activities

Rupak D Sharma writes in The Himalayan Times: [..]If farmers do not get the input on time, agricultural output can dip by 20 to 25 per cent, according to Bhairav Raj Kaini, former director general of Department of Agriculture. A drop in agricultural yield, especially paddy, will hit Nepal’s gross domestic product in the next fiscal year as well, as it accounts for more than a fourth of the country’s total economic output.

[...]The coronavirus crisis has now threatened to eat into this steady income source, as Delhi- to Dubai-based firms are gradually sending Nepali workers, especially those employed in service sector, on unpaid leave or are laying them off. “We have heard news of layoffs and unpaid leaves. But most of the firms in the Gulf and Malaysia, where a big chunk of Nepalis are employed, have not taken such measures for humanitarian reasons. This, however, does not mean there will be no layoffs going forward,” said Kumar Prasad Dahal, director general of the Department of Foreign Employment, adding, “Massive job cuts abroad are inevitable considering the damage the coronavirus crisis has caused to economies.”

This is not a good sign for Nepal that receives around $8 billion in remittances per year, which, as a share, is around a fourth of GDP. A sharp drop in remittance inflow would not only reduce household spending and erode living standard, but also trigger a liquidity crisis.

[...]The unique aspect of the latest economic crisis is that it was not triggered by demand shock but by rapid fall in supply. Production across the globe has dropped or come to a complete halt not because of a slump in demand but because of rapid closure of production units. This has disrupted supply chains and rendered many jobless. This has subsequently forced consumers to tighten their purse strings, triggering a demand shock. This drop in demand may encourage suppliers to further cut back on production leading to more job cuts.

“This is a vicious cycle and is taking the shape of a supply-demand doom loop. This can disrupt economic activities, preventing the economy from reaching its full potential,” said economist Chandan Sapkota. “The country can come out of this precarious macroeconomic situation only if the government launches an effective rescue package that can provide relief to sectors across the board.”

Indian economy to slowdown in FY2021, high unemployment rate, and 1.25 billion workers at risk of losing jobs

SBI pegs FY21 growth at 2.6%, say govt must monetize deficit

From Business Standard: SBI house economists have pegged the growth forecast for January-March at 2.5 per cent and for 2020-21 at 2.6 per cent given the massive disruptions to businesses and the economy due to the COVID-19-driven lockdowns, which has upended at least 70 per cent of the economy.[...]The 21-day lockdown will cost the economy at least Rs 8 trillion, according to a report by SBI Research, which says at least 70 per cent of the economy is on a standstill because of this. 

We estimate another 1.7 per cent impact on real GDP because of the 21-day lockdown in FY21 resulting in at least 70 per cent of the economy at a standstill. We peg FY21 GDP estimate at 2.6 per cent, with a clear downward bias, with Q1 of FY21 GDP numbers witnessing a contraction. FY20 GDP estimates could also see a downward revision from 5 per cent to 4.5 per cent with Q4 growth at 2.5 percent, SBI Research said in a note, adding pegged the total cost of the 21-lockdown at Rs 8.2 lakh crore in nominal terms and output loss at 4 per cent on a conservative approach. But they are quick to add that the economy could rebound if a stronger stimulus is offered.

Given the low market appetite for borrowing, it is imperative that government uses the clause given in FRBM Act and monetize the deficit with the RBI subscribing to the primary issues of the Central government debt and fulfill the supply-demand gap in FY21, the report said. In FY2020, total borrowing by the Centre and states stood at Rs 13.5 lakh crorethe Centre at Rs 7.1 lakh crore and the states combined Rs 6.4 lakh crore.

Given at least estimated 4 per cent slippage in GDP/Rs 8 lakh crore, we expect the Centre and the states could borrow conservatively close to Rs 20 lakh crore in FY21. Thus, it is a must that RBI monetizes the deficit, using the national calamity clause given the stressed market absorption capacity, it says, adding this will add up to 2.5-3 percent of GDP and the government must show it separately as an off-balance sheet item in the budget like a 'COVID bond'.

---

Covid-19 may push 400 mln Indians to poverty: ILO

From Mint: “In India, Nigeria and Brazil, the number of workers in the informal economy affected by the lockdown and other containment measures is substantial. In India, with a share of almost 90% of people working in the informal economy, about 400 million workers are at risk of falling deeper into poverty during the crisis. “Current lockdown measures in India, which are at the high end of the University of Oxford’s COVID-19 Government Response Stringency Index, have impacted these workers significantly, forcing many of them to return to rural areas," it added.

The ILO warning corroborates data from Indian think tanks, which shows how unemployment has tripled in urban India and rural hinterlands within a span of past three weeks. The Centre for Monitoring of Indian Economy has said unemployment rate in India was 23.4% in the week ended 5 April. CMIE data showed that while urban unemployment rate was 30.9%, rural unemployment rate was over 20% and economists have warned that things may only worsen in rural India due to reverse migration in last two weeks.

---

ILO jobs report (COVID-19 ILO Monitor 2nd edition)

According to the new study, 1.25 billion workers are employed in the sectors identified as being at high risk of “drastic and devastating” increases in layoffs and reductions in wages and working hours. Many are in low-paid, low-skilled jobs, where a sudden loss of income is devastating. Worldwide, two billion people work in the informal sector (mostly in emerging and developing economies) and are particularly at risk. The estimates are based on ILO "nowcasting" model (uses real-time economic and labor market data to predict the loss in working hours in Q2 2020).

Large-scale, integrated, policy measures are needed, focusing on four pillars: supporting enterprises, employment and incomes; stimulating the economy and jobs; protecting workers in the workplace; and, using social dialogue between government, workers and employers to find solutions, the study says.

The most highly affected activities include accommodation and food services; real estate, business and administrative activities; manufacturing; and wholesale and retail trade. Medium to highly affected activities include arts, entertainment and recreation and other services; and transport, storage and communication. Medium affected activities include mining and quarrying; financial and insurance activities; and construction. Low affected activities include agriculture, forestry and fishing; utilities; public administration and defence; human health and social work activities; and education. 

The highly affected activities are labor intensive, and employ millions of low-paid and low-skilled workers.