Climate variability impacts on agricultural output in East Africa

Figures & data

Figure 1. Trends in average precipitation and temperature in East Africa.

Table 1. Pesaran CD Test (cross-sectional dependency test)

Part A	CD-test	P-value	Variable	CD-test	P-value
Ln(Output)	29.37	0.00	Ln(Land)	29.70	0.00
Ln(Labour)	28.14	0.00	Ln(Machinery)	26.29	0.00
Ln(Livestock)	23.53	0.00	Ln(Irrigation)	10.60	0.00
Ln(Fertiliser)	6.16	0.00	Ln(Temperature)	30.22	0.00
Ln(Precipitation)	18.25	0.00
Part B: H. M. Pesaran et al. (2008) slope heterogeneity
					P-Value
Model (Equation 2)			$\tilde{\mathrm{\Delta }}$	17.05***	0.000
			${\tilde{\mathrm{\Delta }}}_{Adjusted}$	19.23***	0.000

Table 2. Pesaran (2007) CIPS unit root test results

		A level		First difference
	Intercept	Intercept and trend	Intercept	Intercept and trend
Ln(Output)	−2.472	−2.605	−6.010***	−6.308***
Ln(Machine)	−2.718	−3.178	−5.005***	−5.331***
Ln(Land)	−1.667	−2.080	−5.645***	−5.734***
Ln(Livestock)	−2.326	−2.487	−6.019***	−6.203***
Ln(Labour)	−2.518	−1.049	−0.456***	−0.464***
Ln(Irrigation)	−2.252	−1.970	−5.721***	−5.912***
Ln(Fertiliser)	−3.526	−3.918	−6.190***	−6.420***
Ln(Temperature)	−5.194	−5.207	−6.250***	−6.420***
Ln(Precipitation)	−6.078	−6.309	−5.190***	−6.110***
Temperature variability	−5.103	−5.419	−6.150***	−6.395***
Precipitation Variability	−6.190	−6.420	−6.290***	−6.490***

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, respectively. The critical values of CIPS test at 10%, 5% and 1% levels of significance are −2.21, −2.33 and −2.54 for Intercept, and −2.72, −2.83 and −3.04 for intercept plus trend, respectively.

Table 3. Westerlund And Edgerton (2007) Panel cointegration

	Intercept			Intercept and trend
Statistics	Values	Z-value	P-value	Values	Z-value	P-value
$G_t$	−2.0100	−0.7750	0.2190	−2.0380	−0.8700	0.1920
$G_a$	−11.7***	−2.5340	0.0060	−9.748*	−1.4360	0.0750
$P_t$	−5.678*	−1.3430	0.0900	−5.856*	−1.5220	0.0640
$P_a$	−10.8***	−4.4500	0.0000	−8.9***	−3.1960	0.0010

Note: p < 0.01***, < 0.05** and 0.1*, respectively

Table 4. Pooled mean group (panel ARDL) model test results

Coefficients	Model (1)	Model (2)	Model (3)	Model (4)
Long-run coefficient
Ln(Machinery)	0.087 *** (0.000)	0.091*** (0.000)	0.073 *** (0.004)	0.101*** (0.001)
Ln(livestock)	0.5675*** (0.000)	0.565*** (0.000)	0.655*** (0.000)	0.567*** (0.000)
Ln(Labour)	0.349 *** (0.000)	0.338*** (0.000)	0.321 *** (0.000)	0.326*** (0.000)
Ln(Land)	0.648*** (0.000)	0.607*** (0.000)	0.607*** (0.000)	0.557 (0.000)
Ln(Fertiliser)	0.002 (0.858)	0.006 (0.564)	−0 .013 (0.134)	0.011 (0.356)
Ln(Temperature)	−5.781*** (0.000)	−5.969*** (0.000)	-	0.579* (0.085)
Ln(Precipitation)	-	0.165 (0.233)	-	−5.996 (0.000)
$Variabilit{y}^{Temp}$	−0.095** (0.043)	−0.090* (0.069)	-	-
$Variabilit{y}^{Precip}$	-	-	0.002 (0.311)	0.003 (0.520)
Short-run coefficient
$EC{T}_{\left(t-1\right)}$	−0.208*** (0.002)	−0.199*** (0.002)	−0.197*** (0.000)	−0.181*** (0.002)
$\mathrm{\Delta }Ln\left(Machinery\right)$	0.011 (0.864)	−0.001 (0.994)	0.017 (0.788)	0.004 (0.950)
$\mathrm{\Delta }Ln\left(Livestock\right)$	0.147** (0.046)	0.175** (0.017)	0.154** (0.020)	0.181** (0.011)
$\mathrm{\Delta }Ln\left(Labour\right)$	−1.088 (2.475)	−1.118 (2.492)	−1.064* (2.075)	−1.403 (2.310)
$\mathrm{\Delta }Ln\left(Land\right)$	0.302 (0.553)	0.469 (0.355)	0.389 (0.531)	0.395 (0.579)
$\mathrm{\Delta }Ln\left(Fertiliser\right)$	0.001 (0.916)	−0.001 0.898	−0.001 (0.930)	0.002 (0.888)
$\mathrm{\Delta }Ln\left(Irrigation\right)$	0.025 (0.548)	-	-	-
$\mathrm{\Delta }\mathrm{l}\mathrm{n}\left(Temperature\right)$	−0.223 (0.691)	−0.335 (0.515)	-	−0.200 (0.539)
$\mathrm{\Delta }Ln\left(Precipitation\right)$	-	0.049 (0.289)	-	0.171** (0.067)
$\mathrm{\Delta }Ln\left(aiabilit{y}^{Temp}\right)$	−0.002 (0.88)	−0.005 (0.673)	-	-
Observations	495	495	495	495

Standard errors in parentheses

*** p < 0.01, ** p < 0.05, * p < 0.1

Table 6. Heterogeneous effect of climate variables on agricultural output

Variable	Burundi	Djibouti	Ethiopia	Kenya	Rwanda	Somalia	Sudan	Tanzania	Uganda
$Ln\left(Machinery\right)$	−0.00830	−0.01827	0.10539***	0.06084**	−0.01590	−0.02738	0.07765**	0.05661**	0.14745***
	(0.0240)	(0.02280)	(0.03431)	(0.02604)	(0.03083)	(0.02642)	(0.03517)	(0.02764)	(0.02877)
$Ln\left(Livestock\right)$	0.4114 ***	0.49559***	0.23904*	0.30267***	0.51345***	0.91774***	−0.06107	0.70162***	−0.11228
	(0.09125)	(0.05022)	(0.12870)	(0.08720)	(0.06358)	(0.07472)	(0.05831)	(0.18512)	(0.15489)
$Ln\left(Labour\right)$	0.08556	1.16237***	1.20105***	0.57441***	0.71111***	0.25175***	0.14081***	0.28452 **	−0.04051
	(0.09067)	(0.10866)	(0.07898)	(0.08869)	(0.06957)	(0.04413)	(0.04603)	(0.14328)	(0.14001)
$Ln\left(Land\right)$	−0.08389	0.11931	−0.72601***	−0.04133	0.83716***	−1.70396***	0.56085***	0.57568	2.34100***
	(0.20761)	(0.10253)	(0.18976)	(0.36763)	(0.23458)	(0.38194)	(0.14763)	(0.52084)	(0.59389)
$Ln\left(Fertiliser\right)$	0.04143	0.00393	0.09033***	0.07571***	0.02626	0.03488***	−0.05340	0.01965	0.20574***
	(0.02714)	(0.00818)	(0.03356)	(0.01138)	(0.01681)	(0.01164)	(0.03559)	(0.02065)	(0.01958)
$Ln\left(Precipitation\right)$	−0.27004	0.72475***	−0.03842	−0.05367	1.10233***	0.00410	0.29691	0.39651*	0.07363
	(0.20389)	(0.19013)	(0.09759)	(0.19777)	(0.23225)	(0.19960)	(0.19919)	(0.23725)	(0.28429)
$Ln\left(Temperature\right)$	0.14556	5.03644***	−0.23191	2.02294*	2.76571***	−2.13965**	−1.4054	4.02651***	4.11739***
	(0.40729)	(0.32586)	(0.66974)	(1.07896)	(0.45648)	(1.13533)	(0.80911)	(0.93092)	(1.10957)
$Ln{\left(Variability\right)}^{Precip}$	−0.00234	0.00704**	−0.00897*	−0.00164	0.00989***	0.00078	0.00931	0.00409	0.00707*
	(0.00261)	(0.00324)	(0.00521)	(0.00329)	(0.00324)	(0.00747)	(0.01110)	(0.00301)	(0.00383)
$Ln{\left(Variability\right)}^{Temp}$	−0.02516	0.11009***	−0.09786***	−0.02080	0.04952	0.01055	−0.00474	−0.08516***	0.04907
	(0.03691)	(0.03015)	(0.03286)	(0.03396)	(0.03946)	(0.02934)	(0.03513)	(0.03311)	(0.03954)
Observations	504	504	504	504	504	504	504	504	504
Number of ids	9	9	9	9	9	9	9	9	9

Standard errors in parentheses

*** p < 0.01, ** p < 0.05, * p < 0.1

Table 5. Dynamic fixed effect (DFE) model test results

Coefficients	Model (1)	Model (2)	Model (3)	Model (4)
Long-run coefficient
Ln(Machinery)	− 0.042 (0.067)	−0.049 (0.066)	−0.045 (0.065)	−0.047 (0.066)
Ln(livestock)	0.391*** (0.128)	0.423*** (0.127)	0.396*** (0.115)	0.428*** (0.128)
Ln(Labour)	0.259** (0.111)	0.255*** (0.109)	0.284*** (0.104)	0.280*** (0.104)
Ln(Land)	0.391 (0.352)	0.393 (0.345)	0.408 (0.320)	0.445 (0.344)
Ln(Fertiliser)	−0.013 (0.042)	−0.008 (0.041)	−0.006 (0.042)	−0.003 (0.041)
Ln(Temperature)	−1.286 (4.124)	−0.830 (0.346)	-	0.688 (0.526)
Ln(Precipitation)	-	0.473 (4.054)	-	0.676 (3.594)
$Ln(Variabilit{y}^{Temp})$	−0.115 (0.177)	−0.131 (0.175)	-	-
$Ln\left(Variabilit{y}^{Precip}\right)$	-	-	0.005 (0.008)	0.006 (0.012)
Short-run coefficient
$EC{T}_{\left(t-1\right)}$	−0.109*** (0.020)	−0.111*** (0.019)	−0.109*** (0.0192)	−0.109 (0.0193)
$\mathrm{\Delta }Ln\left(Machinery\right)$	−0.008 (0.022)	−0.005 (0.022)	−0.006 (0.022)	−0.007 (0.021)
$\mathrm{\Delta }Ln\left(Livestock\right)$	0.079* (0.043)	0.077* (0.043)	0.081** (0.044)	0.078 (0.042)
$\mathrm{\Delta }Ln\left(Labour\right)$	−0.010 (0.052)	−0.017 (0.057)	−0.011 (0.054)	−0.019 (0.0549)
$\mathrm{\Delta }Ln\left(Land\right)$	0.377*** (0.136)	0.365*** (0.136)	0.371*** (0.135)	0.370*** (0.136)
$\mathrm{\Delta }Ln\left(Fertiliser\right)$	0.0008 (0.006)	−0.0002 (0.006)	−0.0003 (0.007)	−0.0002 (0.006)
$\mathrm{\Delta }\mathrm{l}\mathrm{n}\left(Temperature\right)$	0.168 (0.428)	0.167 (0.027)	-	0.223 (0.388)
$\mathrm{\Delta }Ln\left(Precipitation\right)$	-	0.044 (0.429)	-	0.071* (0.040)
$\mathrm{\Delta }Ln\left(Variabilit{y}^{Temp}\right)$	−0.006 (0.014)	−0.008 (0.015)	-	-
$\mathrm{\Delta }Ln\left(Variabilit{y}^{Precip}\right)$	-	-	0.0003 (0.001)	−0.0007 (0.0008)
Constant	1.099 (0.369)	−0.799 (0.518)	−0.696* (0.378)	−0.259 (1.089)
Observations	495	495	495	495

Standard errors in parentheses

*** p < 0.01, ** p < 0.05, * p < 0.1

Table A1. Descriptive statistics of economic and climate data

Variables	Mean	St. Dev.
Dependent variable
Output (in millions of international US dollars)	7444.88	5910.00
Independent variables
Physical
Land (1000 persons)	31,226.30	34,763.56
Labour (1000 of hectares)	19,964.49	26,190.84
Machinery	3057.55	3835.89
Livestock (Head count of cattle, sheep and goats, in millions)	28.20	31.30
Irrigation (1000 of hectares)	0.48	0.50
Fertiliser (1000 of hectares)	21,539.26	36,567.19
Climate variables
Temperature	23.83	2.97
Precipitation	64.11	34.65
$Variabilit{y}^{Temp}$	0.06	0.35
$Variabilit{y}^{Precip}$	−0.53	7.30
countries	9
Sample	504

Table A2. Results of the Dumitrescu–Hurlin panel causality test

	O	M	La	Li	Lo	Irr	Fe	T	P	VT	VP
O	–	1.77*	1.95**	2.92***	4.52***	0.83	0.40	0.11	−0.26	0.96	−0.64
		(0.08)	(0.05)	(0.003)	(0.000)	(0.408)	(0.67)	(0.91)	(0.79)	(0.34)	(0.52)
M	5.3***	–	6.6***	4.83***	8.74***	10.***	1.25	0.97	1.32	0.62	0.209
	(0.000)		(0.000)	(0.000)	(0.000)	(0.000)	(0.210)	(0.330)	(0.188)	(0.539)	(0.834)
La	1.52	1.83 *	–	4.02***	6.12***	3.8***	0.12	2.7***	2.40**	−0.13	2.30**
	(0.128)	(0.066)		(0.000)	(0.000)	(0.000)	(0.907)	(0.006)	(0.016)	(0.893)	(0.021)
Li	2.6***	0.43	6.3***	–	3.40***	1.70*	−0.04	3.3***	−0.01	0.82	−0.03
	(0.010)	(0.668)	(0.000)		(0.001)	(0.08)	(0.971)	(0.001)	(0.993)	(0.414)	(0.976)
Lo	23.***	12.7***	10.***	17.35***	–	4.2***	17.***	1.77*	2.93**	0.75	1.42
	(0.000)	(0.000)	(0.000)	(0.000)		(0.000)	(0.000)	(0.076)	(0.003)	(0.455)	(0.155)
Irr	2.85** (0.004)	12.1*** (0.000)	6.9*** (0.000)	1.46 (0.144)	2.69*** (0.007)	–	3.9*** (0.000)	2.36** (0.018)	0.44 (0.663)	−0.92 (0.356)	−0.32 (0.747)
Fe	4.2*** (0.000)	3.01*** (0.003)	0.149 (0.881)	1.92* (0.055)	1.01 (0.311)	0.016 (0.986)	–	0.30 (0.762)	0.68 (0.496)	−0.12 (0.914)	0.24 (0.809)
T	17.*** (0.000)	9.69*** (0.000)	15.*** (0.000)	8.7*** (0.000)	32.4*** (0.000)	14.*** (0.000)	3.6*** (0.000)	–	1.485 (0.137)	24.*** (0.000)	1.262 (0.206)
P	0.734 (0.462)	2.1** (0.034)	1.158 (0.246)	1.063 (0.287)	0.959 (0.337)	0.206 (0.836)	3.9*** (0.000)	0.078 (0.937)	–	−0.827 (0.407)	1.85* (0.064)
VT	5.4*** (0.000)	5.2*** (0.000)	5.4*** (0.000)	4.1*** (0.000)	4.8*** (0.000)	4.9*** (0.000)	1.183 (0.237)	6.8*** (0.000)	−0.749 (0.453)	–	−1.252 (0.210)
VP	0.002* (0.076)	2.5** (0.012)	0.11** (0.028)	−0.90* (0.067)	−0.523 (0.600)	−0.34** (0.030)	2.55** (0.011)	−0.794 (0.427)	−1.403 (0.161)	−1.121 (0.262)	–

Note: P-values are shown in parentheses, ***, **, * represents 1%, 5%, 10% level of significance respectively. O represents output, M (machine), La (land), Li (livestock), Lo (labour),Irr (Irrigation), Fe (fertilizer), T (temperature), P (precipitation), VT (variability in temperature), VP (variability in precipitation)

Pesaran, H. M., Ullah, A., & Yamagata, T. (2008). A bias‐adjusted LM test of error cross‐ section Independence. The Econometrics Journal, 11(1), 105–127. https://doi.org/10.1111/j.1368-423X.2007.00227.x

 Web of Science ®Google Scholar

Pesaran, M. H. (2007). A simple panel unit root test in the presence of cross-section dependence. Applied Econometrics, 22(2), 265–312. https://doi.org/10.1002/jae.951

 Web of Science ®Google Scholar

Westerlund, J., & Edgerton, D. L. (2007). A panel bootstrap cointegration test. Economics Letters, 97(3), 185–190. https://doi.org/10.1016/j.econlet.2007.03.003

 Web of Science ®Google Scholar