^{1}

^{1}

Based on the panel vector autoregressive model (PVAR) and the panel data of China’s 31 provinces from 2010 to 2019, this paper conducts a quantitative analysis on the dynamic balance relationship between China’s express business volume and economic development status in recent 10 years. The results show that: Firstly, there is a long-term equilibrium relationship between the development level of express industry and the economic development level; Secondly, the relationship between them presents regional heterogeneity (that is, the East, Centre and the West show different characteristics); Finally, in the short term, the development of express delivery industry and economic development can well drive each other, but in the long term, this trend is not obvious, and the economic impact of express delivery industry has a certain lag.

Express delivery industry, a new industry that has just developed in the 21^{st} century, is an important part of logistics. It is characterized by its speed, which can deliver goods to the target place in a very short time. In recent years, with the development of long-distance e-commerce, China’s express industry has been developing every rapidly, and has been maintaining a relatively fast development level. In the context of the rapid development of the Internet economy, many buyers and sellers who have never met in the past complete transactions only through E-mail and remote payment, which has promoted the development of the express delivery industry. For exploring the relationship between express delivery industry and economic development, Meng Ran [

Holtz-Eakin (1987) was the first to use the PVAR model to analyze the interaction between endogenous variables of panel data. What he studied was the vector autoregressive model of panel data, which uniformly regarded all variables as endogenous variables and analyzed the relationship between each variable and its lagging items. Using panel data, PVAR model can effectively solve the problem of individual heterogeneity and fully consider the effects of individual and time. The general manifestation of PVAR model is as follows:

y i t = α i , 0 + ∑ j = 1 p α i , j y i , t − j + γ i + θ t + ε i t (1)

where i and t represent region and time, j represents lag period, p stands for lag p period, γ i is individual effect, θ t is time effect, and ε i t is random disturbance term.

In this paper, GDP per capita (unit: 100 million yuan) and express business volume (Q, unit: 10 thousand pieces) were selected, with a time span from 2010 to 2019 (T = 10) and a cross section of 31 provinces, municipalities and autonomous regions in China (N = 31). The sample was selected from China Statistical Yearbook (wap.stats.gov.cn). Descriptive statistical results of the data are shown in

As the panel data used has the property of time series, in order to avoid false regression and false regression, unit root test after taking natural logarithm of the data found that the series was not stable at the 5% test level, so the results were not presented. Therefore, here, ADF test and PP-Fisher test are used to conduct unit root test on variable sum (∆ denoted by first-order difference), and the results are shown in

Because, ΔlnGDP and ΔlnQ are I(1) sequences, the conditions for co-integration

Variable | Observations | Mean | Std. Dev. | Min | Max |
---|---|---|---|---|---|

GDP | 310 | 22,254.03 | 19,254.78 | 512.9 | 107,986.9 |

Q | 310 | 79,378.93 | 189,461.1 | 162.7 | 1,680,594 |

Variables | Methods | ΔlnGDP | ΔlnQ |
---|---|---|---|

All | ADF | −9.88304 (0.0000)** | −3.58041 (0.0000)** |

PP-Fisher | −14.6591 (0.0000)** | −4.97305 (0.0000)** | |

East | ADF | −4.41416 (0.0000)** | −6.49047 (0.0000)** |

PP-Fisher | −6.20695 (0.0000)** | −8.13232 (0.0000)** | |

Centre | ADF | −5.37542 (0.0000)** | −8.38310 (0.0353)** |

PP-Fisher | −1.90823 (0.0000)** | −1.87852 (0.0302)** | |

West | ADF | −4.41285 (0.0000)** | −4.54140 (0.0000)** |

PP-Fisher | −7.47213 (0.0000)** | −5.53729 (0.0000)** |

Note: ∆ indicates first-order difference; **represents a significance level of 5%.

test are met. The first-order difference of variables is stable, which is also the premise of the football co-integration test. The co-integration test of the data is performed, using the Pedroni test, the KAO test, and the Combined Individual test. The results are shown in

Usually, the optimal lag order of the model is determined according to the information criteria AIC, BIC and HQIC. The results are shown in

Granger causality test was further carried out on the ΔlnGDP and ΔlnQ sequence, as shown in

Method | Statistic | Prob |
---|---|---|

Perpdroni | Panel v-Statistic | 4.130149 (0.0000) |

Panel rho-Statistic | −1.849579 (0.0322) | |

Panel PP-Statistic | −4.642120 (0.0000) | |

Panel ADF-Statistic | −3.373187 (0.0000) | |

Group rho-Statistic | 1.515126 (0.9351) | |

Group PP-Statistic | −3.034731 (0.0012) | |

Group ADF-Statistic | −5.520373 (0.0000) | |

Kao | ADF | −3.594079 (0.0002) |

Hypothesized No. of CE(s) | Fisher Stat. * from trace test) | Fisher Stat. * (from_{max} − eigen test) |

None | 489.7 (0.0000) | 444.8 (0.0000) |

At most 1 | 179.6 (0.0000) | 179.6 (0.0000) |

Area | Lag order | AIC | BIC | HQIC |
---|---|---|---|---|

East | 5 | −0.972754* | 0.556865* | −0.390266* |

Centre | 4 | −3.89645 | −2.21501* | −3.35854 |

West | 4 | 0.470693* | 1.99068* | 1.02027* |

H0 | East | Centre | East |
---|---|---|---|

The former sequence is not the Granger cause of the latter | lag = 5 | lag = 4 | lag = 4 |

Statistic | Statistic | Statistic | |

lnGDP-lnQ | 2660.5 (0.0000) | 275.13 (0.0000) | 23.851 (0.0000) |

LnQ-lnGDP | 4174.2 (0.0000) | 400.96 (0.0000) | 88.653 (0.0000) |

Conclusion | Reject | Reject | Reject |

Area | East | Centre | West | ||||
---|---|---|---|---|---|---|---|

s | Variables | lngdp | lnq | lngdp | lnq | lngdp | lnq |

1 | lngdp | 1.000 | 0.000 | 1.000 | 0.000 | 1.000 | 0.000 |

1 | lnq | 0.445 | 0.555 | 0.180 | 0.820 | 0.133 | 0.867 |

2 | lngdp | 0.990 | 0.010 | 0.752 | 0.248 | 0.982 | 0.018 |

2 | lnq | 0.580 | 0.420 | 0.332 | 0.668 | 0.521 | 0.479 |

3 | lngdp | 0.597 | 0.403 | 0.538 | 0.462 | 0.965 | 0.035 |

3 | lnq | 0.483 | 0.517 | 0.343 | 0.657 | 0.771 | 0.229 |

4 | lngdp | 0.641 | 0.3593 | 0.598 | 0.402 | 0.954 | 0.046 |

4 | lnq | 0.483 | 0.517 | 0.344 | 0.656 | 0.747 | 0.253 |

5 | lngdp | 0.874 | 0.126 | 0.601 | 0.399 | 0.956 | 0.044 |

5 | lnq | 0.756 | 0.244 | 0.348 | 0.652 | 0.888 | 0.112 |

6 | lngdp | 0.649 | 0.351 | 0.610 | 0.390 | 0.935 | 0.065 |

6 | lnq | 0.726 | 0.274 | 0.320 | 0.680 | 0.883 | 0.117 |

7 | lngdp | 0.580 | 0.420 | 0.573 | 0.427 | 0.929 | 0.071 |

7 | lnq | 0.776 | 0.224 | 0.323 | 0.677 | 0.886 | 0.114 |

8 | lngdp | 0.733 | 0.267 | 0.554 | 0.446 | 0.931 | 0.069 |

8 | lnq | 0.747 | 0.253 | 0.320 | 0.680 | 0.890 | 0.110 |

9 | lngdp | 0.841 | 0.159 | 0.554 | 0.446 | 0.931 | 0.069 |

9 | lnq | 0.780 | 0.220 | 0.324 | 0.676 | 0.890 | 0.110 |

10 | lngdp | 0.681 | 0.319 | 0.552 | 0.448 | 0.934 | 0.066 |

10 | lnq | 0.758 | 0.242 | 0.331 | 0.669 | 0.892 | 0.108 |

The impulse response measures the short-term effect of one variable on another after a shock of one unit standard deviation.

Variance decomposition is used to analyze the contribution of structural shocks to endogenous variables. It is a deep description and analysis of the interaction and change trend of each variable at present and in the future. As can be seen from

This paper features three economic belt regions on the express industry development and economic development by using the 2010 to 2019 panel data of 31 provinces. The study found that: 1) There is a long-term dynamic equilibrium relationship between them, and they are each other’s granger causality. 2) In the east express industry plays a positive effect on economic growth, but economic development leading role on the development of the eastern region express industry co., LTD. 3) The central and western regions express development impact on economic growth is on the decline, effect is very small, but economic growth has a very good role to express industry development.

Based on this, the following two suggestions are proposed:

1) Strengthen the construction of express industry development system, especially in the eastern region. In the long run, the development of express delivery industry will become a carriage driving economic growth, so it is necessary to speed up the development system of the Courier industry and strengthen its services.

2) Accelerate the pace of economic construction, especially in the western region. Economic development will drive the development of the express delivery industry. Therefore, the central and western regions can promote the development of the express delivery industry by making use of the utilization efficiency of manpower or adjusting the economic structure.

This work is supported by the National Science Foundation of China (No.11561056).

The authors declare no conflicts of interest regarding the publication of this paper.

Shen, S.C. and Shen, J.M. (2021) Research on the Relationship between Chinese Express Delivery Industry and Economic Development Based on PVAR Model. Open Journal of Statistics, 11, 329-336. https://doi.org/10.4236/ojs.2021.112019