Business Cycles

  • Data file(s):
    HardingPagan2002.xlsx HardingPagan2002.dta
    Harding, D., & Pagan, A. (2002), Dissecting the cycle: a methodological investigation, Journal of Monetary Economics, 49, 365-381.
  • Data file(s):
    lusstk.xlsx lusstk.dta
    Pagan, A. R. & Sosunov, K. A., (2003), A Simple Framework for Analyzing Bull and Bear Markets, Journal of Applied Econometrics, 18, 23-46.
  • Data file(s):
    HardingPagan.xlsx HardingPagan.dta
    Harding, D., & Pagan, A. R. (2005), A Suggested Framework for Classifying the Modes of Cycle Research, Journal of Applied Econometrics, 20, 151-159.


  • Data file(s):
    DungeyPaganSVAR.xlsx DungeyPaganSVAR.dta
    Dungey, M., & Pagan, A. (2000). A structural VAR model of the Australian economy. Economic Record, 76, 321-342.

    Mardi Dungey writes about the genesis and development of this series of papers as follows:

    "This paper came about because I had a practical background in international economics and the Australian economy when I came to my PhD, and am very obsessed with data quality so I think Adrian was convinced we had a chance of getting something to work. Read more

    At the time VARs had been around for a little while but the existing models were all closed economy — not very suitable for Australia. So Adrian and I started to try to work out how to do an Australian model. I had to work out how to even code this thing — I had an old version of Gauss, with a manual for an even older version and of course there were no such things as internet search engines. Coding it was hard, and I was massively inefficient. At the end of my PhD I was the proud possessor of the best desktop in the department in order to be able to bootstrap those confidence bands — and even then we had to do them in batches!! Putting together the empirical model was really challenging. We did a LOT of experimenting — there was no blueprint from anyone else — and my thesis contains chapter after chapter of alternative specification, and a separate volume on the coding!

    The big breakthrough came when we decided that we would put a delay in the effect of monetary policy shocks on the economy — suddenly we got rid of the price puzzle plague. (We also spent considerable time on UIP restrictions but we never resolved that one.) In all the work I have done subsequent on VAR in Australia I have still not found a way to remove the delay on monetary policy restriction without price puzzle reappearing, although I know others disagree with that. Cushman and Zha (1997) was happening at the same time — I remember when we first discovered their parallel work — they were definitely faster to print than us but the work was happening at the same time.

    I don't think I realised as a PhD student that this paper would have such an impact. I was so excited to get a revise and resubmit from Economic Record. But of course the changes the reviews prompted were horrendously time consuming. We had to completely respecify it to put in an investment channel — which was a good suggestion and did improve the model. But it took a long time. In meantime I had finished my PhD, taken up my first academic job at La Trobe, and had my first child. I remember ringing Adrian when I was pregnant with the second and telling him this paper had to go back before that child was born – and it did!

    Adrian and I were both aware that there were lots of issues arising from the model. We have been working away on these both separately and together over the past decade and a half. Adrian has a done a lot of work on theoretical and data coherence in these models — including his important papers with Hashem Pesaran and recently with Lance Fisher. Some of that material was incorporated into the paper he and I wrote published in Economic Record in 2000. I have tended to concentrate on incorporating the methodological developments in the literature into applied empirical work, including work with Renee Fry on NZ and most recently Denise Osborn on the EU and US. But there is still a lot more to do!"

    Professor Mardi Dungey Biography

Nonlinear Time Series

  • Data file(s):
    AndersonAthanasopoulosVahid_G7.xls AndersonAthanasopoulosVahid_G7.dta
    Anderson, H.M., Anthansopoulous, G and Vahid, F. (2007) Nonlinear autoregressive leading indicator models of output in G7 countries. Journal of Applied Econometrics, 22, 63-87.
  • Data file(s):
    ausu.xlsx ausu.dta
    Bardsen G, Hurn A S and McHugh Z (2011) Asymmetric unemployment dynamics Studies in Nonlinear Dynamics and Econometrics, 16, 1-22.

Phillips Curve

  • Data file(s):
    GPT Data.xls GruenPaganThompson.dta
    Gruen, D., Pagan, A., & Thompson, C. (1999). The Phillips curve in Australia. Journal of Monetary Economics, 44, 223-258.

    Adrian Pagan writes about the origins of the paper:

    "Bob King asked me to write about the Phillips Curve for Australia as he knew that Phillips had done one when he came to Australia for a visit around 1958. Read more

    Apart from the UK that was his only work on the curve. In it he had to allow for the fact that wages were effectively indexed by the Arbitration Commission i.e. real rather than nominal wages. In the UK he had tried a real wage form by putting import prices in the equation (these represented food prices) but this turned out to be insignificant.

    I knew David Gruen well and as he was research director of the RBA and had a Phillips Curve in the small model he was using to forecast with at the Bank. He became involved because we wanted to see how the Phillips Curve had affected policy here. This issue became a bigger part of the paper after we started going through the minutes of the Board. Christopher Thompson was a graduate who was assigned to do the quantitative work by David. I wrote the Kalman filter stuff but he cleaned it up and also cleaned up the data.

    As we went alone we changed specifications and I put in the Markov Switching stuff at the end, as I was doing a lot of similar modeling for the business cycle and it seemed to be a way of allowing the gap to be a latent variable in a Phillips Curve. I don't think anyone read that even though I think it was novel. The paper has had a pretty good level of cites so I think JME are probably happy about publishing it."

  • Data file(s):
    GPT Data.xls GruenPaganThompson.dta
    Enders, W., Hurn, S., (2002), Asymmetric Price Adjustment and the Phillips Curve, Journal of Macroeconomics, 24, 395-412.

    Stan Hurn writes:

    "I first met Walt Enders when he visited the University of Glasgow for a year in 1990. We did some collaborative research on exchange rate behaviour which proved to be very productive and enjoyable. Read more

    As a result we stayed in touch after Walt's return to Iowa State. In 2000, after my appointment at QUT and about the time when Walt was deciding on a move from Iowa State to the University of Alabama (a move he subsequently made), I invited him to come out to Brisbane for a 3 month visit so that we could renew our research links. Walt had recently been on sabbatical at UCSD was thinking a lot about nonlinear models following his work with Clive Granger on asymmetric adjustment and unit roots (JBES 1998) and with Pierre Siklos on threshold cointegration (JASA, 2001). As I had been looking at the Gruen, Pagan and Thompson (JME, 1998) paper with some interest, it seemed quite natural to re-examine the Australian Phillips curve to see if a nonlinear specification fitted the data. We purposefully decided to use the same data as the JME paper to facilitate drawing comparisons."

  • Data file(s):
    auswp.xlsx auswp.dta
    Bardsen G, Hurn A S and McHugh Z (2007) Modelling wages and prices in Australia, Economic Record, 83, 143-158.