Agriculture Reference
In-Depth Information
of sample, or from different years from the years in which the prices are being predicted. The
results provide insight into the critical interaction between global prices and growing conditions
in the region where a market obtains its food grains. The local growing conditions and the
ability of traders to access additional grain from the international market interact to affect local
food prices. Thus when determining the impact of local growing seasons on food prices, con-
sideration of the international price of a comparable grain is important. Particularly in regions
where information on the availability and cost of transportation and the local connections with
the international market are unknown, this insight is important (Kshirsagar 2012).
Local market response to shocks
The modeling framework presented above will allow analysts to categorize markets with
regard to the likely impact of droughts and other weather shocks on food prices. Some
markets have more food that is sourced locally than others, and every market has different
transportation and transaction costs that are difficult to quantify from afar. These characteriza-
tions are important for the kind of qualitative assessments of the impact of food price variabil-
ity on food security that are routinely done by FEWS NET and other organizations that
monitor food security. Local weather events, even those that do not actually result in produc-
tion declines, can have an impact on food prices locally. The lack of data on specific market-
level factors such as trade volumes, informal transactions costs and other unobserved,
time-varying market heterogeneity significantly reduces the ability of food security analysts to
assess what these impacts are or the likely future change of food prices due to a shock. Thus
using the observable weather information along with global food prices in a quantitative
model can improve an organization's ability to estimate food price changes in regions where
food security is a concern.
What impact do weather shocks have on any one country or region? To explore this issue,
we will return to our example in Niger, where there are many markets with food price obser-
vations over three decades. A study by Essam (2013) showed that millet markets in Niger in
zones with excess millet production tend to generate leading price signals to periphery markets
located in less intensive production zones. However, the temporal nature of these relation-
ships is varied and recent trends suggest that there have been improvements in overall market
integration in Niger. A correlation analysis conducted by Essam shows that in years following
weather-related production shocks, market prices tend to move more closely in tandem than
in years of abundant production. It is only through an analysis of the spatial relationships
between markets that permit such a conclusion, although analyzing connection between
markets is very difficult without significant information on trading networks and transaction
costs that are rarely available in real time (Essam, 2013).
To further investigate this latter conclusion, Essam estimated a price dispersion model that
analyzes the effect of exogenous weather shocks as measured by NDVI on price spreads
between markets. This price dispersion modeling results suggest that negative (positive) NDVI
shocks decrease (increase) price dispersion. As the extent of an NDVI shock grows, captured
by the percent of markets with a negative NDVI shock, absolute price dispersion declines
between six and ten CFA per kilogram of millet (Essam, 2013). These results are robust across
both standard ixed-effects models and specifications that include a dynamic adjustment factor,
with adjustments made to the standard errors to account for general forms of cross-sectional
and temporal dependence.
Search WWH ::
Custom Search