Did the TERS policy save jobs during the COVID-19 pandemic?

South Africa’s Temporary Employer-Employee Relief Scheme (TERS) has arguably served as the country’s most important labour market intervention in response to the COVID-19 pandemic to date. As the government winds down the policy two years after its inception, a key question is: was it successful in achieving its primary aim of saving jobs?

Introduction

 Job retention policy has been a primary measure used by governments around the world to mitigate job losses in response to the COVID-19 pandemic. Broadly speaking, these policies seek to help employers retain workers and avoid the potentially costly process of hiring and training new workers as economic activity recovers, as well as to help workers avoid adverse labour market scarring effects associated with periods of unemployment.[1] These policies typically take one of two forms: short-time work schemes (which subsidise hours not worked) or wage-subsidy schemes (which subsidise hours worked, but which can also be used to top up wages). The latter has been particularly prevalent during the pandemic, with 60% of countries having implemented some form of wage subsidy by January 2022.[2]

 In South Africa, where unemployment is amongst the highest globally, such a job-retention policy has arguably served as one of the government’s most important economic interventions during the pandemic to date: the Temporary Employer-Employee Relief Scheme (TERS). The TERS is a wage subsidy that has provided support to employers who fully or partially closed their operations in response to lockdown regulations. Simply put, the policy aimed to save jobs by subsidising firms’ labour costs, with benefits ranging between R3 500 and R6 730 per worker per month – much larger than the government’s social grant expansions during the pandemic period.

Two years have passed since the inception of the TERS. By April 2022, the policy had benefited 5.7 million workers (61 – 70% of the formal, private employed population in 2020[3]) at a cost of R64 billion.[4] Now that the government is preparing to wind down the policy, the question is whether it succeeded in achieving its primary aim of saving jobs? To our knowledge little to no research has been conducted in this regard, but in our recently-published journal article, we seek to provide some answers. We show that the TERS benefited millions of workers, and that receipt was highest during the most stringent lockdown levels. We document significant changes in the recipient population over time and show that the distribution of receipt and benefits were relatively distribution-neutral. Although not causal, we find evidence that TERS receipt is associated with an 18.1 percentage point increase in the probability of job retention, but only during the initial ‘hard’ lockdown period. Our findings highlight the potential of wage subsidies to mitigate job loss and should be kept in mind when considering how to provide targeted support to workers during future crises.

The development and evolution of the TERS

The initial TERS Directive was issued on 25 March 2020. The policy was administered by the Unemployment Insurance Fund (UIF), which allowed government to leverage existing infrastructure to implement the policy swiftly and effectively.[5] The policy targeted workers who remained employed but had suffered income loss because of full or partial closure of their employer’s operations due to the pandemic. Initially, however, only UIF-contributing workers were eligible. Although this included most workers (8.5 million, or 61% of employment in 2020Q1[6]) given the labour market’s well-documented formal sector concentration, it excluded workers in the informal sector as well as UIF non-contributors in the formal sector. Following legal challenges, at the end of May 2020 eligibility was expanded to include any worker who could prove an existing employment relationship.

 At first, workers did not apply for and receive benefits directly. Generally, the employer or the relevant bargaining council applied on a worker’s behalf. This was to decrease administrative burdens related to processing large application volumes. After receiving payment, employers were then liable to pay the benefits to their workers within two days and submit proof of payment to the UIF within five days. To mitigate the risk of benefits being trapped at the employer or bargaining council level, the UIF published a list of recipient employers in the public domain so that workers could follow up and submit queries. Later, payments were made directly into workers’ bank accounts, although applications still had to be submitted by employers. Despite initial pay-out delays, most benefits (96%) were paid within 30 days after application receipt.[7] Throughout the remainder of 2020 and 2021, the TERS was subject to various extensions and amendments following varying lockdown regulations. The timeline in Figure 1 below provides a summary of these changes.

Figure 1: Timeline of the TERS in the context of South Africa’s national lockdown

Authors’ own arrangement. Source: Köhler and Hill (2022).

 TERS benefits are calculated according to the usual UIF benefit formula as laid out in the Unemployment Insurance Act. Briefly, an eligible worker’s benefit depended on their calculated Income Replacement Rate which itself depended on their salary or wage. Simply put, benefits ranged progressively between 38% and 60% of a worker’s usual wage but could not fall below the national minimum wage (NMW) of R3 500 per month or go above R6 730 per month, regardless of if the calculated benefit fell above or below these amounts. This meant that sub-minimum wage workers were entitled to benefits exceeding their usual wage. Overall, this benefit formula was progressive, with lower-wage workers receiving larger benefits in relative terms. Figure 2 presents a visual summary of how benefits are calculated. As an example, a worker earning R9 125 per month is entitled to an IRR of 43%, resulting in a monthly benefit of nearly R3 900.

 Figure 2: Simulation of the calculation of TERS benefits

Authors’ own arrangement. Source: Köhler and Hill (2022).

Data and methodology

 In our study, we use individual-level microdata from all fives waves of the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDS-CRAM) conducted between May 2020 and May 2021. The NIDS-CRAM is a broadly representative, longitudinal telephonic survey designed as a ‘barometer’ for assessing the socio-economic impact of the COVID-19 pandemic on South African individuals and households.[8]

 No other representative household survey includes data on TERS receipt. The survey asked every employed adult whether they received a TERS benefit in a specific reference month.[9] We conducted a three-part analysis. First, we analysed aggregate trends in TERS receipt over time. Second, we considered how receipt has varied between different groups of workers over time. Third, we analysed the relationship between receipt and job retention by making use of multivariate regression and propensity score matching techniques. This essentially allows us to compare the job retention probabilities of workers who did and did not receive the TERS but otherwise resemble each other in terms of observable characteristics like age, race, sex, occupation, and province. Importantly, this technique does not allow us to rule out any differences in unobservable characteristics (i.e. characteristics we don’t have data on) between recipients and non-recipients. Therefore, we can identify only a correlational relationship (“Are TERS recipients more likely to remain employed?”) and not a causal one (“Does TERS receipt increase the probability of remaining employed?”). Interested readers are referred to our paper for more detail on our data and method.

Who benefitted from the TERS?

We estimate that over four million unique workers received the TERS at least once in our five reference months. As shown in Figure 3, the reach of the TERS was highest during the beginning of the lockdown.^[10]In April 2020 during lockdown level 5 alone, the policy reached 1.8 million workers (14% of all workers), and a similar level in June during level 3. Over time, just over half (53%) of the recipients in April 2020 continued to receive TERS in June 2020. Throughout the remainder of 2020 and the beginning of 2021 many workers continued to benefit, but far fewer than before. This is likely due, at least in part, to payment delays and backlogs, but additionally to recovering economic activity that reduced the number of eligible workers, as well as changes to the policy’s eligibility criteria. Since only the employed were eligible, some degree of this variation in receipt over time can also, of course, be explained by job loss.

 

Figure 3: Trends in aggregate TERS receipt, April 2020 – March 2021

Authors’ own arrangement. Source: Köhler and Hill (2022).

 Some groups benefitted more than others. As reported in Table 1, men, African/Black individuals, those living in urban areas, those with a matric qualification or less, those aged 35-59 years, and those working in semi-skilled jobs in the tertiary (services) sector accounted for most recipients. Notably, workers who were unable to work from home represented up to 94% of TERS recipients. Importantly, given that only the employed were eligible for the TERS, these differences in receipt shares are at least partially explained by differences in employment shares. In our paper, we find that even after accounting for this, several groups continued to disproportionately benefit, including men, semi-skilled workers, workers in the secondary sector, and those who could not work from home.

 How did TERS receipt vary across the wage distribution? TERS receipt was relatively distribution-neutral (neither pro-poor nor pro-rich) over the period, but among recipients, the distribution of benefits was slightly regressive: for instance, in several months, approximately 40% of benefits accrued to the top 30% of earners. This latter finding is likely simply a function of the design of the TERS benefit formula, which ensures benefits are higher in absolute terms for higher-wage workers. Despite this, it is worth noting again that lower-wage workers received higher benefits in relative terms. We estimate that the average TERS benefit among the poorest 20% of earners was equivalent to about seven times the group’s average pre-pandemic wage, compared with a ratio of 0.6 for the richest 40% of earners. 

 Table 1: Share of TERS recipients (%), by demographic and labour market characteristic

		Wave 1 (April 2020)	Wave 2 (June 2020)	Wave 3 (October 2020)	Wave 4 (January 2021)	Wave 5 (March 2021)
Gender	Male	61.0	60.5	63.2	65.5	63.8
	Female	39.0	39.5	36.8	34.5	36.2
Race	African/Black	74.8	69.2	76.9	79.2	90.9
	Coloured	13.1	10.3	14.1	8.6	4.7
	Asian/Indian	2.0	5.9	1.0	7.7	1.0
	White	10.0	14.6	8.0	4.5	3.4
Area	Traditional	10.8	10.2	11.9	12.8	9.8
	Urban	81.6	87.6	86.8	85.7	85.6
	Farms	7.5	2.3	1.3	1.5	4.6
Skill level	High-skilled	16.0	16.4	11.7	11.6	10.0
	Semi-skilled	68.0	63.3	60.7	50.0	64.4
	Less-skilled	16.0	20.3	27.6	38.4	25.6
Sector	Primary	14.9	8.6	10.3	4.1	4.9
	Secondary	28.3	31.0	20.1	40.1	29.6
	Tertiary	56.8	60.4	69.6	55.8	65.5
Written contract	No	19.1	24.5	19.3	27.2	15.0
	Yes	80.9	75.5	80.7	72.8	85.0
Education	Up to primary	8.3	11.1	10.6	3.8	20.5
	Incomplete secondary	28.8	35.3	26.9	46.1	30.3
	Complete secondary	28.6	22.0	25.8	16.4	19.6
	Tertiary	34.3	31.7	36.6	33.7	29.6
Age (years)	18-34	42.1	40.8	40.8	40.4	35.3
	35-59	54.6	52.7	54.9	59.1	63.2
	60+	3.2	6.5	4.3	0.4	1.5
Work from home	No	0.0	81.1	81.2	91.2	93.9
	Sometimes	0.0	11.9	13.5	5.3	3.1
	Yes	0.0	7.1	5.3	3.5	3.1

Authors’ own arrangement. Source: Köhler and Hill (2022).

Did the TERS save jobs?

Was the TERS successful in mitigating job loss? In our analysis, we indeed find statistically significant evidence of a positive relationship between TERS receipt and job retention, however only during the most stringent lockdown period. Specifically, for workers with the same observed characteristics, we find that TERS receipt in April 2020 was associated with an 18.1 percentage point increase in the probability of remaining employed in the same job in June 2020. Importantly, although this finding is consistent with the policy being successful, as discussed above our data and method here do not allow us to identify a causal effect.

Interestingly, we find no evidence of any relationship between receipt and job retention for the remainder of 2020 or beginning of 2021. Why might this be the case? We discuss several explanations in our paper, such as varying eligibility criteria of the policy over time; or the fact that more organised firms might have been more likely to apply for TERS in early periods. Moreover, firms may only have used the TERS in the initial period to retain productive workers who, in the absence of receipt, would have been retained anyway. Overall, though, our findings remain consistent with the notion that the TERS had a positive effect on job retention, but only during the initial ‘hard’ lockdown period.

Conclusion

 South Africa’s Temporary Employer-Employee Relief Scheme (TERS) has arguably served as the country’s most important labour market intervention in response to the COVID-19 pandemic to date. In a country characterised by extreme unemployment, job retention has been as a key welfare objective during this time. Millions of workers have benefited from the policy since its inception. Two years later, as the government winds it down, a key question is whether the policy was successful in achieving its primary aim of saving jobs.

 Our study shows that the TERS reached over four million workers at least once between April 2020 and March 2021, with receipt being highest during the most stringent lockdown levels. There were some significant changes in the TERS recipient population over time. Notably, workers from several vulnerable groups experienced greater-than-proportional coverage than would be expected given their overall shares of employment, such as less-skilled workers, those with lower levels of formal education, and those unable to work from home. By making use of a multivariate regression and propensity score matching technique, we find that TERS receipt in April 2020 is associated with an 18.1 percentage point increase in the probability of remaining employed in the same job in June 2020. However, we find no such evidence for the remainder of the period.

 The TERS provided an important source of income relief for many vulnerable firms and workers, and although we cannot infer causality from our results, they are consistent with the policy having a positive effect on job retention during the most stringent period of South Africa’s lockdown. As an important next step, we are currently conducting more research on the TERS with the aim of establishing causality. Overall, these results point to the role of wage subsidies in the mitigation of job loss and should be kept in mind when considering how to provide targeted support to workers during future crises in the country. However, policymakers ought to also consider how discrepancies between claim and payment periods may jeopardize the welfare of vulnerable workers by denying or delaying their support, ultimately threatening the recovery of the labour market in general.

References

 Auditor-General South Africa (AGSA), 2020a. First special report on the financial management of government’s Covid-19 initiatives. AGSA, Pretoria.

 Auditor-General South Africa (AGSA), 2020b. Second special report on the financial management of government’s Covid-19 initiatives. AGSA, Pretoria.

 Gentilini, U, Almenfi, M, Iyengar, H, Okamura, Y, Downes, JA, Dale, P, Weber, M, Newhouse, D, Alas, CR, Kamran, M, Mujica, IV, Fontenez, MB, Ezzat, M, Asieduah, S, Martinez, VRM, Hartley, GJR, Demarco, G, Abels, M, Zafar, U, Urteaga, ER, Valleriani, G, Muhindo, JV & Aziz, S, 2022. Social protection and jobs responses to COVID-19: A real-time review of country measures. Version 16. World Bank, Washington, DC.

 Giupponi, G & Landais, C, 2020. Building effective short-time work schemes for the COVID-19 crisis, VoxEU.org, 01 April.

Gronbach, L, Seekings, J & Megannon, V. 2022. Social Protection in the COVID-19 Pandemic: Lessons from South Africa. CGD Policy Paper 252. Washington, DC: Center for Global Development.

 Ingle, K, Brophy, T & Daniels, R, 2021. NIDS-CRAM Panel user manual. NIDS-CRAM Technical Document C1.

 Keenan, E & Lydon, R, 2020. Wage subsidies and job retention. Central Bank of Ireland Economic Letter 2020(11), 1–12.

Köhler, T & Hill, R, 2022. Wage subsidies and COVID-19: The distribution and dynamics of South Africa’s TERS policy. Development Southern Africa. DOI: 10.1080/0376835X.2022.2057927.

National Income Dynamics Study-Coronavirus Rapid Mobile Survey (NIDS-CRAM) 2020, Waves 1, 2, 3, 4, 5 [datasets]. Version 2.0.0. Cape Town: Allan Gray Orbis Foundation [funding agency]. Cape Town: Southern Africa Labour and Development Research Unit [implementer], 2020. Cape Town: DataFirst [distributor], 2020, 2021.

Nxesi, T. 2022. Two-year anniversary of the UIF’s Covid-19 TERS – Thulas Nxesi, MP. Minister of Employment and Labour. 04 April 2022.

OECD (Organisation for Economic Co-Operation And Development), 2020. Job retention schemes during the COVID-19 lockdown and beyond.

Statistics South Africa (StatsSA), 2020a. Quarterly Labour Force Survey 2020: Q1; Q2; Q3; Q4  [datasets]. Version 1. Pretoria: Statistics South Africa [producer]. Cape Town: DataFirst [distributor].

 [1] Giupponi & Landais, 2020; Keenan & Lydon, 2020; OECD, 2020

[2] Gentilini et al., 2022.

[3] Own calculations using Statistics South Africa’s QLFS for all four quarters of 2020.

[4] Nxesi, 2022

[5] Gronbach et al, 2022

[6] Own calculations using Statistics South Africa’s QLFS for 2020Q1.

[7] AGSA, 2020a; 2020b

[8] See Ingle et al (2021) for more information on the NIDS-CRAM sampling design.

[9] Given this, our estimates on receipt represent workers who received the TERS in a particular month but not necessarily for a particular month. It is important to note then that our estimates are likely to underestimate the true reach of the programme given delays between application and payment.

[10] This data is correct up to March 2021. These figures may have changed due to eligibility changes in the policy after March 2021.

Download article