MODELING THE IMPACT OF CLIMATE VARIABILITY ON LOCAL FOOD PRICES - Food Security, Food Prices and Climate Variability - page 151

its (relatively) higher value to weight ratio. Consequently, it is profitable to import rice, while
arbitrage opportunities for coarse grain trade are significantly lower. Further, several countries
(many in West Africa) consume rice, and need to import rice to meet domestic consumption
needs (Moseley et al ., 2010). Second, world maize prices do influence local price formation
in Latin America, and thus these countries' maize prices are affected by international price
shocks. These countries are more open and have better transport infrastructure than those in
Sub-Saharan Africa. Finally, world wheat prices also influence domestic prices in parts of
Eastern Europe and Central Asia as well as in South Asia. Table 7.3 shows the results of a
model run using maize prices in non-African countries that shows the influence of inter-
national maize prices or domestic NDVI increases.

Locations affected both by world prices and by local weather shocks

In the maize results provided in Table 7.3 , we showed that weather and price shocks often
interact. Although in many of the model results we do not see a big improvement in the
market functioning for Maize in Africa when including international prices, there are a few
markets where the two interact significantly. These locations are typically well-connected
food surplus locations in countries that trade internationally. Many commodity-market pairs
show a slight improvement when NDVI and international price shocks are used together.
When we use all data available and explore the impact both of weather and of international
shocks on price dynamics, we find that the relationship of local food prices with shocks can
vary within a region and within a country. Plate 22 shows a map that categorizes markets into
those that are affected by both weather and international price shocks, and those that are not.
West African markets in Mali, Niger and Burkina are not affected by international prices (and
thus are not flagged as being affected by both) because of their relative isolation from inter-
national markets. East African markets, however, are affected by both kinds of shocks and are
particularly sensitive to local weather dynamics. The South Asian markets in Afghanistan,
Pakistan and India also show sensitivity to both kinds of shocks. When the interaction between
local and global shocks occurs at the same time, households can experience extremely intense
food security stress (Garg et al ., 2013; Bradbear and Friel, 2013).

TABLE 7.3 Model errors and percent increase of maize prices and NDVI values

Country Global price increase (%) Anomalous NDVI

increase (%)

Forecast RMSPE (%)

Kenya 0.11 (0.27) -1.29** (0.65) 6.3

Uganda 0.36 (0.46) -2.32* (1.21) 26.2

Mexico 0.33 *** (0.12) -0.45 (0.39) 10.4

Cameroon -0.09 (0.18) 0.05 (0.70) 4.7

Guatemala 0.23** (0.1) 0.43 (1.24) 6.6

Chad 0.21 (0.27) -3.2** (1.5) 28.9

Note
Significance is denoted with *** , percent change is given in parentheses.