Global slack hypothesis: evidence from China, India and Pakistan

Zaniboni (2008) also derives the CPI-based NKPC using the relationship between the terms of trade and the exchange rate.

Different lag length measures (AIC, SBIC and HQIC) are used to determine the order of lag length.

It is a common practice in estimating NKPC to use lagged values of inflation, output gap and other additional variables, e.g., interest rate spread, as instruments (Gali and Gertler 1999). These variables are relevant for forecasting expected inflation, an endogenous variable, in the first-stage regression. I also apply a variety of instruments for each country to ensure that the estimated parameters are consistent and stable. Initially, four lags of the discount rate (end of period) are used for India and Pakistan. The size and statistical significance of estimates are not affected while the Hansen’s (1982) test does not pass the validity condition for instruments. The interest rate spread (lending rate minus deposit rate) and deposit interest rate with four lags are also used for China. In general, instruments are relevant and valid based on Hansen (1982) test and Stock and Yogo (2005) test. The results are also robust across different sets of instruments.

The theoretical models in Eqs. (1)–(2) relate domestic goods price inflation to domestic output gap and foreign output gap, while Eq. (4) relate the CPI inflation to the domestic and foreign output gaps. As a result, I use the domestic goods inflation and CPI inflation in estimation, accordingly. As these theoretical models relate individual inflation process of a country to its respective domestic output gap and foreign output gap, the panel data analysis would not be strictly necessary in estimation. As a result, the theoretical models are estimated country-by-country basis.

The ten largest trading partners of China in 2010 are Australia, Brazil, Hong Kong, Germany, India, Japan, Republic of Korea, Malaysia, Singapore and the USA. The ten largest trading partners of India are Belgium, Hong Kong, China, Germany, Netherlands, Saudi Arabia, Singapore, United Arab Emirates, the UK and the USA. The ten largest trading partners of Pakistan are Afghanistan, China, India, Iran, Japan, Kuwait, Malaysia, Saudi Arabia, United Arab Emirates and the USA.

The endogeneity problem is resolved when a structural model (i.e., Phillips curve equation, Euler equation and monetary policy rule) is estimated as used by Milani (2009). In this case, the domestic output gap (and also foreign output gap) is modelled in the framework, which does not require to lag the trade weight.

The structural breaks may affect the reduced form parameters of the closed-economy NKPC (Zhang et al. 2008).

The lag length criteria (AIC, SBIC and HQIC) are used to determine order of lags for inflation. The AIC and HQIC lag length criteria suggest two lags for inflation, while SBIC suggests one inflation lag for China. For Pakistan, the AIC, HQIC and SBIC lag length criteria suggest one lag for inflation, while zero lag for inflation in the case of India.

For China, the data are available from 1979 and results are reported over the full sample (1979–2009) and the sub-sample (1985–2009) periods.

These results are available upon request.

These results are not reported due to the space limitations but are available upon request.