Do not quote or cite without permission

Electricity

Access to electricity increased from 52 to 78 percent between 2005/6 and 2014, with much of the improvement coming from what appears to be a significant expansion of rural electrification (Table 10). Electricity coverage in rural areas increased from 36 to 68 percent during this period. It is not possible to infer from the data whether the improvement was due to the expansion of access to the national grid, which was a major infrastructure challenge for the country, or due to other small and medium sources including solar and wind power.

Table 10: Access to electricity, 2005/6 and 2014

	2005/6	2014
Urban	95.3	99.0
Rural	35.5	68.2
Poorest quintile	24.6	49.9
2	41.9	72.9
3	51.2	80.8
4	62.4	88.1
Richest quintile	79.3	95.6
Total	51.9	77.5

Source: World Bank staff calculations based on HBS 2005/6 and HBS 2014.

Education

School enrollment rates increased between 2005/6 and 2014, and fewer children were out of school (Table 11). Gross enrollment rates increased from 73.4 to 88.3 percent, while net enrollment rates increased even faster, from 66.4 to 84.5 percent. The proportion of children out of school more than halved, from 33.6 to 15.5 percent. This progress was particularly pronounced at the lower end of the distribution. Gender gaps in all three of these indicators were reduced significantly, although these were no eliminated entirely.

Table 11: Gross and net enrollment in school for children, percent

	Gross enrollment		Net enrollment		Not in school
	2005/6	2014	2005/6	2014	2005/6	2014
Urban	87.2	91.2	81.5	87	18.5	13
Rural	68.1	87	61.6	83.5	38.4	16.5
Male	82	90.1	75.9	86.2	24.1	13.8
Female	62.6	86.1	56	82.4	44	17.6
Poorest quintile	59.7	87.8	53.3	85.3	46.7	14.7
2	71.2	86.6	65.6	83.2	34.4	16.8
3	74.2	86.9	67.5	82.9	32.5	17.1
4	77.8	89.3	71.3	85.1	28.7	14.9
Richest quintile	87	91	80.7	86.3	19.3	13.7
Total	73.4	88.3	66.4	84.5	33.6	15.5

Source: World Bank staff calculations based on HBS 2005/6 and HBS 2014.

Education quality, especially in public schools, is often a key issue in many developing countries. Private alternatives have emerged in many settings and are available to parents able to afford these options. In Yemen too, enrollment in private schools increased from 4.1 percent in 2005/6 to 7.4 percent in 2014. And, as one may speculate, the use of private alternatives for children’s education was more prevalent among the better off parents (Figure 13). Data from 2014 show that around 18 percent of children from the richest quintile went to private schools, while the number for those in the poorest quintile was less than 1 percent. The steep jump from 6.1 percent to around 18 percent between the 4^th to the 5^th quintile suggests that private schools cater to the richest consumption expenditue quintile group in Yemen. Nonetheless, the fact that even in the richest segment fewer than one-fifth of children used private providers suggests that public schools remain still extremely important for the country.

Figure 13: Use of private schools for children's education

Source: World Bank staff calculations based on HBS 2005/6 and HBS 2014.

Health

Households with higher consumption self-reported sickness or accidents at a higher level than poorer households (Table 12). Around 24 percent of the household in the top 60 percent reported being sick or having an accident in 2014. This was in contrast to the bottom 40 percent among whom only 20 percent reported being sick or having an accident. It is often difficult to infer anything from this statistic on the health shocks experienced by households across the income ladder. For the same kind of health shock, health seeking behavior increases with increasing living standards, so it is not surprising that self-reports of being in need of medical attention is higher among the wealthy.

But when they needed care, the wealthy were also more likely to receive health care, with those in the top 60 percent having a 10-percentage-point higher likelihood of receiving care when they needed it than the bottom 40 percent. Poorer households were more likely to use public facilities for their health care, although the difference was only slight and the overall level of public-facility use declined for the entire population. There was a higher likelihood of utilizing a facility located in the neighborhood for the richer households. One possible reason for this is that richer people tend to live in urban areas, as noted earlier.

Table 12: Illness and health care, 2005/6 and 2014

	2005/6			2014
	Top 60%	Bottom 40%	Total	Top 60%	Bottom 40%	Total
Had sickness or accident	10.1	9.4	9.9	23.7	20.1	22.3
Received medical care	76.4	66.9	72.8	79.9	69.1	76
Went to a public facility	30.6	33.2	31.5	21.7	25.9	23.1
Located in the neighborhood	29.9	24.9	28.2	40.6	26.5	36.1

Source: World Bank staff calculations based on HBS 2005/6 and HBS 2014.

There were some interesting differences in the burden of diseases between rich and poor in 2014 (Figure 14). The rich and the poor had almost equal probabilities of having an accident or catching pneumonia, or even to some extent being afflicted with diarrhea or other intestinal diseases. However, the non-poor were more likely to report ENT and other flu-related diseases. Looking more deeply into incidence by quintile, it becomes apparent that the likelihood of reporting malaria/fever had a clear and a montonic wealth gradient, whereas the ENT and flu diseases appeared fairly even for the bottom four quintiles and spiked up considerably in the top 20 percent (Figure 15). While there were no further data here to adequately explain this curious pattern, two conjectures can be made. First, perhaps there were physical environmental factors associated with malarial incidence that were correlated with place of residence, which may in turn be correlated with income. In other words, poorer households may live in areas with higher susceptibility to certain diseases. Second, for ENT and flu-type issues, the wealthiest may be more likely to seek medical attention than any other groups for the same kind of symptoms.

Figure 14: Types of disease and accident, 2014

Source: World Bank staff calculations based on HBS 2014.
Figure 15: Types of disease and accident, 2014

Source: World Bank staff calculations based on HBS 2014.

Vaccinations are linked with reduction in infectious diseases worldwide and the WHO considers a child to be fully immunizaed only after the child receives one BCG shot, at least three polio shots, three DPTs and one shot for measles. It is also recommended that children receive these vaccines within one year of birth. Data from the most recent round of the HBS (2014) show that Yemen was very far from universalizing full immunization of children in their second year: only two in five Yemeni children were found to be fully immunized, with important variations in vaccination rates by specific vaccines (Table 13). While the results of the 2014 HBS and the Demographic and Health Survey, which was conducted in 2013, are very similar, the immunization results in the Multi Indicator Cluster Survey (MICS) and the earlier round of the HBS, both conducted broadly in 2005 and 2006, appear to diverge dramatically. In particular, the numbers from the earlier HBS stand in sharp contrast to reports of an improvement in overall immnization rates between 1997 and 2006 reported in the DHS 2013 report.¹⁰ Due to these methodological issues, this note refrains from commenting on any recent trends in Yemen and instead uses 2014 HBS to analyze the vaccination rates across different groups and regions (Table 14).

Table 13: Vaccination rates for children (12-23 months)

	BCG	DPT 3	Polio 3	Measles	Full vaccination	No vaccination
YMICS 2006	69.0	61.0	63.0	65.0	38.0	12.0
HBS 2005/6	74.3	67.6	78.9	75.7	59.8	7.6
YDHS 2013	67.6	59.6	58.7	63.3	42.6	16
HBS 2014	74.1	48.8	67.8	71.8	41.5	8.9

Note: Data for HBS 2005/06 and HBS 2014 were calculated by World Bank staff based on HBS 2005/6 and HBS 2014. Data for YDHS 2013 and YMICS 2006 are taken from the Yemen DHS 2013 report.

Poorer children were less likely to be vaccinated. In 2014, only 36 percent of children living in poor households were fully vaccinated, in contrast to 48 percent for the children living in non-poor households. There was also a significant difference between children living in urban and rural areas, with children in urban areas having a higher rate of full immunization. The gap in full immunization was more than 20 percentage points. There was no major difference in children receiving no vaccinations at all by urban or rural region.

Table 14: Vaccination rates for children (12-23 Months), 2014

	Full vaccination	No vaccination
Non-poor	47.9	7.5
Poor	35.8	10.1
Urban	57.7	7.4
Rural	35.4	9.4
Poorest quintile	35.9	13
2	37.8	7.8
3	33.9	7.4
4	48.3	9.7
Richest quintile	55.5	6.5
Total	41.5	8.9

Source: World Bank staff calculations based on HBS 2014.