Purpose

World over, change of base year in the gross domestic product (GDP) is a standard practice of GDP estimation. However, unless a consistent series of GDP is released with respect to the new base for the earlier period, the existence of multiple growth rates creates problems for applied researchers, policymakers and the general public alike. Faced with such a menu of GDP series researchers often try to interpolate a consistent series of GDP. The main purpose of this paper is to analyses the nature of the data generating process of such multiple interpolated series of quarterly growth rates and tries to discern the consistency of such processes.

Design/methodology/approach

The present paper tries to look into the statistical implications and complications of such interpolated quarterly GDP/growth series in India in terms of three series of GDP, namely, with 1999–2000, 2004–2005 and 2011–2012 as its bases.

Findings

The analysis reveals that as a result of a change of base year, the nature of the data generating process of the old and new GDP series could undergo changes and experience different breakpoints. While all these conclusions seem to be valid for GDP growth at quarterly intervals, taking the data at annual frequency is less problematic.

Practical implications

The observation suggests that in most applied work, researchers may not have the luxury of only working with annual data and certain consistency checks will be necessary to check the veracity of the results based on quarterly data with those based on annual data. Second, moving forward it may be useful for the Authorities to make a transition to a chain-based linking method rather than fixed time-period-based bases as is currently done.

Originality/value

The analysis of Indian GDP in this paper is, perhaps, indicative of the fact that usage of quarterly GDP data is to be handled with caution and it is preferable that any serious empirical analysis uses annual GDP data whenever it is available/feasible. The comparison of GDP growth rates at different frequencies and examining the true nature of the process are quite unique in their contribution towards empirical macroeconomic research.

中文翻译：

---

基准年和印度 GDP 增长的变化：不可知论

目的

在世界各地，国内生产总值 (GDP) 基准年的变化是 GDP 估算的标准做法。然而，除非针对较早时期的新基数发布一致的 GDP 序列，否则多重增长率的存在会给应用研究人员、政策制定者和一般公众带来问题。面对这样的 GDP 序列菜单，研究人员经常尝试插入一致的 GDP 序列。本文的主要目的是分析这种多个插值系列季度增长率的数据生成过程的性质，并试图辨别这些过程的一致性。

设计/方法/方法

本文试图从三个 GDP 序列（即以 1999-2000 年、2004-2005 年和 2011-2012 年为基础）来研究印度这种内插的季度 GDP/增长序列的统计意义和复杂性。

发现

分析表明，由于基准年的变化，新旧GDP序列的数据生成过程的性质可能会发生变化并经历不同的断点。虽然所有这些结论似乎都适用于每季度的 GDP 增长，但以年度频率获取数据问题不大。

实际影响

观察表明，在大多数应用工作中，研究人员可能无法只使用年度数据，因此需要进行某些一致性检查，以检查基于季度数据的结果与基于年度数据的结果的准确性。其次，向前推进，当局过渡到基于链的链接方法而不是像目前所做的基于固定时间段的方法可能是有用的。

原创性/价值

本文中对印度 GDP 的分析可能表明，应谨慎处理季度 GDP 数据的使用，任何严肃的实证分析都最好在可用/可行的情况下使用年度 GDP 数据。不同频率下 GDP 增长率的比较和对这一过程的真实性质的检验，在它们对实证宏观经济研究的贡献方面是非常独特的。