K. Ludi: consumption behaviour in Zambia
12
are inaccurate and biased. Thus, the ECM is used via its residuals to adjust the long-run
coefficients and their corresponding t-statistics.
The residuals from the ECM are regressed on the variables included in the long-run
equation multiplied by the negative coefficient of the residuals from the cointegrating
equation retrieved from the ECM. The results of the Engle-Yoo regression are shown in
table 5.
Table 5: Results of third-step Engle-Yoo regression (dependent variable: RES_ECM)
Variable |
Coefficient |
Std. Error |
0.421942*LN_GDP_ZK |
0.000253 |
0.001416 |
....................0C421942*R^lENDrAiE.................... |
...............0.000232................ |
..................0.000601................... |
The values from this regression are used to adjust the long-run coefficients, as shown in
table 6.
Table 6: Adjustment of long-run cointegrating coefficients and t-statistics
Variable |
Coefficient calculation |
Adjusted |
Standard error |
Adjusted t- |
LN_GDP_ZK |
0.999958 + 0.000253 |
1.000211 |
0.001416 |
706.364 |
................r_lendrate................ |
...............-0.002285"'+"0.000232............... |
........................-0.002053........................ |
...........................0.000601................. |
3.416..................... |
Now that the cointegrating coefficients have been adjusted, they may be interpreted with
confidence. Since the t-statistics are statistically significant at a 1 per cent level of
significance,12 it can be claimed that real short-term lending rates and national output
influence PCE in the long run. Economically, the coefficient of LN_GDP_ZK represents
the marginal propensity to consume,13 and is thus a very interesting result, due to the fact
that it is larger than 1. This implies that Zambians are spending more on consumption that
what they effectively earn. This result however should not be interpreted too stringently,
11
12
13
Calculation of the adjusted t-statistic: (adjusted coefficient / std. error).
The t-statistics are larger than the calculated critical value of 2.75 (df = 31).
Calculated as the first order derivate of PCE with respect to income.
More intriguing information
1. Ultrametric Distance in Syntax2. Segmentación en la era de la globalización: ¿Cómo encontrar un segmento nuevo de mercado?
3. Response speeds of direct and securitized real estate to shocks in the fundamentals
4. The name is absent
5. The name is absent
6. WP 36 - Women's Preferences or Delineated Policies? The development or part-time work in the Netherlands, Germany and the United Kingdom
7. Micro-strategies of Contextualization Cross-national Transfer of Socially Responsible Investment
8. ASSESSMENT OF MARKET RISK IN HOG PRODUCTION USING VALUE-AT-RISK AND EXTREME VALUE THEORY
9. Cancer-related electronic support groups as navigation-aids: Overcoming geographic barriers
10. The name is absent