Sampling methodology and fieldwork changes in the October Household Surveys and Labour Force Surveys

Figures & data

Figure 1: Distribution of households per EA in OHS 1995–OHS 1998 and OHS 1999–LFS September 2007

Table 1: Single-person households, average household size and multiple household dwelling units

Survey wave	Estimated number of single-person households	Average household size (weighted)	Estimated number of single-person households from the census	Percentage of dwelling units with more than one household reported (unweighted) (%)	Percentage of households at multiple household dwelling units (weighted) (%)
OHS 1994	999 504	4.49
OHS 1995	1 141 363	4.45		2	2.4
OHS 1996	1 063 911	4.44	Census 1996: 1 725 765	1.3	1.4
OHS 1997	1 287 104	4.40
OHS 1998	1 626 600	4.19
OHS 1999	2 088 252	4.00		4.0	8.9
LFS 2000:1	2 353 726	3.90		5.7	13.1
LFS 2000:2	2 415 377	3.87		5.1	11.3
LFS 2001:1	2 541 079	3.81		7.1	16.5
LFS 2001:2	2 564 525	3.83	Census 2001: 2 479 259	7.7	16.6
LFS 2002:1	2 750 219	3.77		9.0	19.2
LFS 2002:2	2 602 528	3.82		3.2	7.1
LFS 2003:1	2 771 083	3.73		3.5	7.6
LFS 2003:2	2 861 516	3.69		3.9	8.4
LFS 2004:1	2 919 406	3.66		3.4	7.6
LFS 2004:2	2 822 535	3.63		4.9	10.4
LFS 2005:1	2 834 873	3.65		6.1	13.2
LFS 2005:2	2 863 295	3.64		6.4	14.5
LFS 2006:1	2 743 018	3.69		5.8	12.7
LFS 2006:2	2 952 101	3.61		6.2	13.8
LFS 2007:1	3 065 676	3.55		6.6	15.2
LFS 2007:2	2 947 933	3.60		4.3	10.6

Note: OHS and LFS data weighted using cross-entropy weights discussed in Branson & Wittenberg (2014). 10% census samples weighted using Stats SA household weights.

Table 2: Household weight regressions

	OHS 1995	OHS 1996	OHS 1997	OHS 1998
Multiple household dummy	5.75	–
	(4.95)
Backyard dummy			−1.393	−5.455**
			(1.609)	(2.362)
Constant	305.6167***	569.523	310.544***	489.422***
	(0.52)		(0.197)	(0.292)
N	29 700	15 921	29 811	18 981
R^2	0.747	1	0.945	0.976

Notes: The dependent variable is the household weight released by Stats SA. Standard errors in parentheses. Cluster fixed effects included in all regressions. The variation in weights in OHS 1996 is fully explained by the cluster fixed effects; that is, weights are constant within enumeration areas.

**Significance at the 5% level.

***Significance at the 1% level.

Table 3: Employment in the OHSs and LFSs by gender and for single-person households

Survey	Total employment			Employment in single-person households
	Total	Female	Male	Total	Female	Male
OHS 1994	9 635 642	3 679 058	5 956 585	644 114	67 186	576 927
OHS 1995	10 296 924	3 855 119	6 441 806	860 598	106 158	754 440
OHS 1996	9 537 055	3 634 004	5 903 051	671 088	102 874	568 214
OHS 1997	9 643 678	3 637 833	6 005 845	837 025	118 309	718 716
OHS 1998	10 204 242	3 947 211	6 245 568	1 089 430	174 701	914 274
OHS 1999	11 192 713	4 606 829	6 578 508	1 456 399	339 866	1 116 533
LFS 2000:2	12 777 248	5 537 203	7 239 362	1 774 785	399 092	1 375 693
LFS 2001:2	11 783 379	4 992 034	6 791 345	1 800 188	484 460	1 315 728
LFS 2002:2	11 880 779	4 988 898	6 887 766	1 753 604	500 008	1 253 596
LFS 2003:2	11 963 947	5 056 975	6 906 972	1 961 255	510 772	1 450 483
LFS 2004:2	12 290 751	5 151 728	7 133 709	1 867 264	488 329	1 378 935
LFS 2005:2	13 082 463	5 538 361	7 539 253	1 977 816	475 534	1 501 441
LFS 2006:2	13 528 143	5 750 893	7 776 861	2 054 500	514 941	1 539 559
LFS 2007:2	14 031 007	6 009 158	8 013 690	2 077 648	547 832	1 529 817

Note: Cross-entropy weights used to weight the data; see Branson & Wittenberg (2014).

Branson, N & Wittenberg, M, 2014. Reweighting South African national household survey data to create a consistent series over time: A cross-entropy estimation approach. South African Journal of Economics 82(1), 19–38. doi: 10.1111/saje.12017

 Web of Science ®Google Scholar