(5 %). The number of shocks reported during this period amounted to 2.4 on
average, and this figure did not differ significantly between the wealth groups (cf.
Table 5.2). However, there is evidence to suggest from the results that the nature of
shocks experienced was related to households’ poverty status (chi-square test
significant at p < 0.01). In particular, animal deaths were cited as a shock by
poor households more than non-poor households (30 % vs. 22 %), and due to the
fact that the poor tend to live at higher altitudes and on steeper terrain, they
experienced floods less frequently than the non-poor (13 % vs. 18 %), but landslides
more frequently (8 % vs. 1 %).
As Table 5.2 shows, the shocks experienced between 2005 and 2011 caused an
estimated average total loss of 26 million VND per household, not taking into
account the mitigating effects of any coping measures that may have been applied
(see below). About two-thirds of this loss on average was attributable to a drought
that occurred in 2010. The data indicates that shock-induced losses were significantly greater for wealthier households, both in their entirety and as a specific result
of the 2010 drought. This can be explained by differences in land productivity, that
is, the wealthier tended to attain higher yields and gross margins when cultivating
the main crop – maize, due to higher levels of input use and possibly superior crop
management practices, leading to greater losses when this one crop failed.
Regarding the coping strategies applied by farmers, the data indicate that in 53 %
of the shock events, households did not apply any coping measures that could be
perceived as such. In 30 % of the shock events households drew upon their own
savings, in 10 % of the events they sold livestock and in 9 % of the events they
borrowed money from friends or relatives. We did not find any significant
differences between households below or above the poverty line, but again, the
relative poverty classification revealed some evidence of a relationship between
household wealth status and the primary coping measures applied (chi-square test
significant at p < 0.05). While in only 17 % of events the poorest tercile used their
own monetary savings to cope with the shock, the wealthiest tercile did so in 28 %
of cases, and as a consequence, the poorest tercile utilized consumption loans taken
from informal sources more often. Another difference concerned the use of temporary off-farm employment as a coping strategy, which over the study period was
used in 6 % of the shock events experienced by the poorest tercile, but only in 2 %
of cases by the wealthiest tercile. We did not find however, that households from
different wealth strata used different strategies to cope with the 2010 drought.
Concerning the incidence of shock-induced consumption reductions, which is
our measure of households’ resilience against shocks, again only the relative
poverty classification revealed statistically significant evidence of a higher level
of resilience among wealthier households. Considering all the shocks experienced,
a reduction in household consumption levels occurred among approximately 60 %
of the wealthiest tercile, as compared to 64 % and 70 % in the median and poorest
terciles respectively. When looking at the 2010 drought in particular, 65 % of
households in the wealthiest tercile had to reduce their level of consumption, as
opposed to more than 80 % for the other two terciles. We can therefore conclude
that the majority of households lack a reasonable level of resilience against shocks,
with this share being somewhat higher among the poorest tercile households.
186
C. Saint-Macary et al.
average, and this figure did not differ significantly between the wealth groups (cf.
Table 5.2). However, there is evidence to suggest from the results that the nature of
shocks experienced was related to households’ poverty status (chi-square test
significant at p < 0.01). In particular, animal deaths were cited as a shock by
poor households more than non-poor households (30 % vs. 22 %), and due to the
fact that the poor tend to live at higher altitudes and on steeper terrain, they
experienced floods less frequently than the non-poor (13 % vs. 18 %), but landslides
more frequently (8 % vs. 1 %).
As Table 5.2 shows, the shocks experienced between 2005 and 2011 caused an
estimated average total loss of 26 million VND per household, not taking into
account the mitigating effects of any coping measures that may have been applied
(see below). About two-thirds of this loss on average was attributable to a drought
that occurred in 2010. The data indicates that shock-induced losses were significantly greater for wealthier households, both in their entirety and as a specific result
of the 2010 drought. This can be explained by differences in land productivity, that
is, the wealthier tended to attain higher yields and gross margins when cultivating
the main crop – maize, due to higher levels of input use and possibly superior crop
management practices, leading to greater losses when this one crop failed.
Regarding the coping strategies applied by farmers, the data indicate that in 53 %
of the shock events, households did not apply any coping measures that could be
perceived as such. In 30 % of the shock events households drew upon their own
savings, in 10 % of the events they sold livestock and in 9 % of the events they
borrowed money from friends or relatives. We did not find any significant
differences between households below or above the poverty line, but again, the
relative poverty classification revealed some evidence of a relationship between
household wealth status and the primary coping measures applied (chi-square test
significant at p < 0.05). While in only 17 % of events the poorest tercile used their
own monetary savings to cope with the shock, the wealthiest tercile did so in 28 %
of cases, and as a consequence, the poorest tercile utilized consumption loans taken
from informal sources more often. Another difference concerned the use of temporary off-farm employment as a coping strategy, which over the study period was
used in 6 % of the shock events experienced by the poorest tercile, but only in 2 %
of cases by the wealthiest tercile. We did not find however, that households from
different wealth strata used different strategies to cope with the 2010 drought.
Concerning the incidence of shock-induced consumption reductions, which is
our measure of households’ resilience against shocks, again only the relative
poverty classification revealed statistically significant evidence of a higher level
of resilience among wealthier households. Considering all the shocks experienced,
a reduction in household consumption levels occurred among approximately 60 %
of the wealthiest tercile, as compared to 64 % and 70 % in the median and poorest
terciles respectively. When looking at the 2010 drought in particular, 65 % of
households in the wealthiest tercile had to reduce their level of consumption, as
opposed to more than 80 % for the other two terciles. We can therefore conclude
that the majority of households lack a reasonable level of resilience against shocks,
with this share being somewhat higher among the poorest tercile households.
186
C. Saint-Macary et al.