Skip to main content
Journal for Labour Market Research

Table 4 Model estimates of lockdown effects on formal sector employment probabilities.

From: Lockdown stringency and employment formality: evidence from the COVID-19 pandemic in South Africa

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

 

POLS

POLS

POLS

FE

DiD (POLS)

DiD (POLS)

DiD (FE)

Treatment

 − 0.078***

 − 0.079***

 − 0.062***

 

 − 0.084***

 − 0.068***

 
 

(0.009)

(0.009)

(0.007)

 

(0.010)

(0.008)

 

Post

 

 − 0.079***

 − 0.050***

 − 0.055***

 − 0.082***

 − 0.055***

 − 0.053***

  

(0.004)

(0.004)

(0.004)

(0.005)

(0.005)

(0.005)

Treatment × post

    

0.009

0.011

 − 0.004

     

(0.008)

(0.008)

(0.008)

Time-varying controls

N

N

Y

Y

N

Y

Y

Time-invariant controls

N

N

Y

N

N

Y

N

Individual FE

N

N

N

Y

N

N

Y

Constant

0.535***

0.582***

 − 0.637***

1.365**

0.584***

 − 0.634***

1.361**

 

(0.006)

(0.006)

(0.051)

(0.571)

(0.006)

(0.050)

(0.571)

Observations

26,286

26,286

26,069

26,069

26,286

26,069

26,069

R2

0.006

0.012

0.308

0.869

0.012

0.308

0.869

  1. This table presents estimates of specification (1) with a binary formal sector employment variable serving as the dependent variable. Sample restricted to the working-age population (15–64 years) as of 2020Q1. POLS = pooled ordinary least squares, FE = fixed effects, DiD = Difference-in-Differences. Standard errors presented in parentheses and are clustered at the panel level. All estimates weighted using sampling weights. Time-varying controls include age, highest education level, and employment type. Time-invariant controls include sex, racial population group, province of residence, a binary urban residence indicator, an ‘industry or occupation job-mover’ indicator, and a ‘formality job mover’ indicator as described in Sect. 3.3
  2. ***p < 0.01, **p < 0.05, *p < 0.10
Back to article page