28.01.2015 Views

The Future of Smallholder Farming in Eastern Africa - Uganda ...

The Future of Smallholder Farming in Eastern Africa - Uganda ...

The Future of Smallholder Farming in Eastern Africa - Uganda ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Table 6a--Unit root tests for price series <strong>in</strong> the preliberalization era<br />

Market p Criterion ADF test<br />

statistic<br />

Null hypothesis (unit<br />

root present)<br />

Kiritiri (no trend) 0 AIC, SBC, HQC -1.563 Cannot reject<br />

Kiritiri (with trend) 0 AIC, SBC, HQC -1.680 Cannot reject<br />

Siakago (no trend) 0 AIC, SBC, HQC -1.422 Cannot reject<br />

Siakago (with trend) 0 AIC, SBC, HQC -1.505 Cannot reject<br />

Kitale (no trend) 0 AIC, SBC, HQC -2.144 Cannot reject<br />

Kitale (with trend) 0 SBC, HQC -2.110 Cannot reject<br />

Nairobi (no trend) 0 AIC, SBC, HQC 0.026 Cannot reject<br />

Nairobi (with trend) 1 AIC, SBC, HQC -1.965 Cannot reject<br />

Migori (no trend) 0 SBC, HQC -1.645 Cannot reject<br />

Migori (with trend) 2 AIC, SBC, HQC -2.833 Cannot reject<br />

Awendo (no trend) 0 AIC, SBC, HQC -0.933 Cannot reject<br />

Awendo (with trend) 0 AIC, SBC, HQC -2.235 Cannot reject<br />

95% critical value for the ADF statistic (no trend) -2.997<br />

95% critical value for the ADF statistic (with -3.662<br />

trend)<br />

<strong>The</strong> next step <strong>in</strong>volved differenc<strong>in</strong>g the price series and repeat<strong>in</strong>g the unit root test on the<br />

price differences. <strong>The</strong> equations estimated were:<br />

=<br />

i, α + 0 δ 1 P + m<br />

i, t − 1 ∑ = n<br />

+<br />

= α m Δ m Pi, t −m<br />

U (6)<br />

t<br />

Δ 2 P t<br />

Δ 2 P =<br />

i, t α + 0 δ 1 P t<br />

(7)<br />

i 1<br />

m 1<br />

m n<br />

m 1<br />

, − + ∑ = =<br />

+ α m Δ m P +<br />

i, t −m<br />

β 1 T U t<br />

<strong>The</strong> model selection criteria suggest orders <strong>of</strong> augmentation <strong>of</strong> between zero and two with<br />

most regressions be<strong>in</strong>g <strong>of</strong> order zero. In all cases, and regardless <strong>of</strong> the order <strong>of</strong><br />

augmentation chosen, the computed ADF statistics are well above the 95 percent critical<br />

value <strong>of</strong> the test given at the last two rows <strong>of</strong> Tables 7a and 7b. Hence the hypothesis that the<br />

first differences <strong>of</strong> the price series have unit roots is firmly rejected. From the evidence<br />

provided by this test, it was concluded that the price series are <strong>in</strong>tegrated <strong>of</strong> order one--that is,<br />

all price series are I(1). Once the order <strong>of</strong> economic <strong>in</strong>tegration was established, the next step<br />

was to test for co<strong>in</strong>tegration <strong>of</strong> the price series.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!