Uzbekistan Quality Job Creation as a Cornerstone for Sustainable Economic Growth

Trade and Foreign Direct Investment in Uzbekistan
201
and is statistically significant, which confirms the results of a series of other 
empirical studies (Head and Mayer 2013, Kang and Dagli 2018). 
Finally, the GDP and distance are combined with other measures to estimate 
the export determinants of Uzbekistan. We use the extended gravity model
6
similar to Kang and Dagli (2018), which takes the following log-linear form: 
lnX
it
=
 
β
0
+
β
1
 lndist
i
 
+
 
β
2
 lnGDP
t
 
+
 
β
3
 lnPOP
it
+
β
4
 C
i
 
+
 
β
5
 lnRER
it
+
φ
i
+
 
γ
t
+
ε
it
, 
where 
ln
denotes log, 
X
it
is the value of deflated exports of Uzbekistan to its 
trading partner 
i
at year 
t
, 
dist
i
is the population-weighted distance between 
Uzbekistan and its importers, and 
GDP
t
is the value of deflated GDP of 
Uzbekistan. 
POP
it
is the population of partner destinations that vary over time.
C
i 
are the common control variables that include dummies
7
for geographical 
contiguity, membership in the EU, and WTO membership. 
RER
it
is the bilateral 
real exchange rate, calculated by the consumer price index weighted nominal 
exchange rate (Kang and Dagli 2018). 
φ
i
and 
γ
t
are importer and year fixed 
effects, respectively. Together they control for endogeneity arising from 
omitted variables, such as tariff and nontariff barriers. With the inclusion of 
importer fixed effects, the initial GDP variables of importing countries drop 
out. Finally, 
ε
it
is the error term. The dataset on control variables as well as the 
population and distance were extracted from CEPII. The nominal exchange 
rates were retrieved from the Thomson Reuters DataStream platform, 
and consumer price indexes were retrieved from the World Bank’s World 
Development Indicators. The data cover 47 importers and 17 years (2000–
2016), generating 799 observations.
In practice, the estimated parameters of a gravity equation are interpreted 
in terms of elasticities. For example, the distance parameter (in logarithms) 
is the elasticity of trade to distance, reflecting the percentage variation in 
trade following a 1% increase in distance. The distance parameter typically 
serves as a proxy for trade costs. Intuitively, the longer the bilateral distance, 
the higher the associated trade costs. A geographical contiguity variable in 
the analysis is used to capture information costs. In our dataset the countries 
include dummies for Kazakhstan, the Kyrgyz Republic, and Tajikistan, which 
share common cultural and language features. Such features help to decrease 
information costs because firms share similar business environments. The 
dummies for membership in the EU and the WTO reflect the importer’s 
6
The naive form of the gravity equation includes the GDP of country of origin and country of destination, and 
their physical distance. 
7
Dummies are variables that take a value of 0 or 1. For example, a dummy takes the value 1 if an importing 
country is a member of the WTO, and 0 otherwise. 

---