Testing Panel Data Regression Models with Spatial Error Correlation



rook or a queen type weight matrix, and the rows of this matrix are standardized so that they
sum to one.1 We fix
¾X + ¾° = 20 and let ½ = ¾X=(¾X + ¾°) vary over the set (0; 0:2; 0:5; 0:8).
The spatial autocorrelation factor
λ is varied over a positive range from 0 to 0:9 by increments
of
0:1. Two values for N = 25 and 49, and two values for T = 3 and 7 are chosen. In total,
this amounts to
320 experiments.2 For each experiment, the joint, conditional and marginal
LM and LR tests are computed and
2000 replications are performed. In a first draft of this
paper we reported the two-sided LM and LR test results to show how misleading the results
of these tests can be. These results are available upon request from the authors. In this
version, we focus on the one-sided version of these tests except for testing
H0: ¾X = 0 where
we thought a warning should be given to applied econometricans using packages that still
report two-sided versions of this test.

3.1 Joint Tests for H0a: λ = ¾2 = 0
O

Table 1 gives the frequency of rejections at the 5% level for the handy one-sided Honda-type
LM test statistic
LMH given in (2.18), the GHM test statistic given in (2.19) and LRJ given
in (2.23). The results are reported for
N = 25; 49 and T = 3; 7 for both the Queen and Rook
weight matrices based on
2000 replications. Table 1 shows that at the 5% level, the size of
the joint LR test
(LRJ) is not significantly different from 0:05 for all values of N and T and
choice of the weight matrix
W . The same is true for LMH and GHM except for N = 25 and
T = 3 where they are undersized. The power of all three tests is reasonably high as long as
A > 0:3 or ½ > 0:2. In fact, for ½ 0:5 this power is almost one in all cases. For a fixed A or
½, this power improves as N or T increase.

1 The weight matrix with first-order contiguity according to the rook criterion has the cells immediately
above, below, to the right, and to the left, for a total of four neighboring cells. The weight matrix with first
order contiguity according to the queen criterion is eight cells immediately surrounding the central cell, see
Anselin and Rey (1991).

2The Monte Carlo experiments were also run for negative λ ranging between -0.1 and -0.9. The results
were similar and are not reproduced here to save space.

10



More intriguing information

1. THE DIGITAL DIVIDE: COMPUTER USE, BASIC SKILLS AND EMPLOYMENT
2. Indirect Effects of Pesticide Regulation and the Food Quality Protection Act
3. Motivations, Values and Emotions: Three Sides of the same Coin
4. The name is absent
5. The name is absent
6. The name is absent
7. Protocol for Past BP: a randomised controlled trial of different blood pressure targets for people with a history of stroke of transient ischaemic attack (TIA) in primary care
8. Regulation of the Electricity Industry in Bolivia: Its Impact on Access to the Poor, Prices and Quality
9. ALTERNATIVE TRADE POLICIES
10. Evidence on the Determinants of Foreign Direct Investment: The Case of Three European Regions
11. The name is absent
12. Deletion of a mycobacterial gene encoding a reductase leads to an altered cell wall containing β-oxo-mycolic acid analogues, and the accumulation of long-chain ketones related to mycolic acids
13. The name is absent
14. The name is absent
15. Towards a framework for critical citizenship education
16. Pursuit of Competitive Advantages for Entrepreneurship: Development of Enterprise as a Learning Organization. International and Russian Experience
17. The name is absent
18. Who runs the IFIs?
19. Fiscal Insurance and Debt Management in OECD Economies
20. Retirement and the Poverty of the Elderly in Portugal