Project Aims

We are a team of academic researchers from Massey University partnering with NZ businesses and government agencies in the quest to develop AI & machine learning-based solutions to estimating key economic indicators like GDP in real-time using data from multiple and diverse sources.

There is a real interest and need to estimate how the economy is performing right now. What we know about the national GDP is never current and up-to-date. We may know what the GDP for the previous quarter was, but we can only find this out with a lag of several months. In the case of Regional GDP, these figures are made available only once a year and thus our grasp of regional economic wellbeing right now is not accessible.

This project is therefore unique in its undertaking. Our aims are to embrace the seismic shifts our societies are undergoing in respect to the availability of new and rich data sources that can be mined and converted into valuable insights. We have created a GDP Dashboard that predicts the GDP value for today both national and for all the regions, in addition to the prediction of contribution to the GDP that each insustry sector is currently making.

A number of institutions already provide GDP forecasts. None use the same data or the machine learning algorithms employed in this project. As this project matures, acquiring additional datasets and improved prediction accuracies, we believe that this project can serve government, businesses, media and other institutions in providing a useful estimate of economic activity that can complement and augment currently used data.

Given enough meaningful data and appropriate algorithms, it is only matter of time before a robust and accurate real-time estimate of national GDP is created. We hope that this project will lay the foundations for this system and to that end we invite other institutions, businesses and government agencies to partner with us in helping to achieve this goal.

How GDP works

GDP is used as a measure for comparing how well, or poorly, a country is faring compared to other countries. The idea of measuring the whole economy was first developed in 1652 by William Petty who used it to estimate England and Wales’ ability to fund a war, to be financed through taxation. The concept of measuring a country’s “national income” was developed through the eighteenth century working with the premise that it depended on how much was available for spending now, and how much must be retained for growing the national asset base.

In 1776, Adam Smith challenged the definition of national income by only counting labour that contributed to making physical commodities, agriculture and industry. Adam’s view was that the provision of services was actually a cost to the economy and did not add any value to the underlying asset – it was an unproductive activity. This view predominated throughout the nineteenth century until the advent of neoclassical economists like Alfred Marshall rejected the distinction between goods and services. This was just as well, as by the late 1980s, services would make up two thirds of Western economies output.

The Great Depression in the 1930s provided the impetus to accurately measure economic output as a way to end the United Kingdom’s unprecedented economic slump. This was mirrored in the United States, where Simon Kuznets, who later won the Nobel Memorial Prize in Economic Science for his efforts, produced a report for US Congress in 1934, which showed that the national income had halved between 1929 and 1932. During the 1930s the national income, and statistics in general, were useful again as a means of working out how to fund the looming Second World War. John Maynard Keynes pamphlet "How to pay for the War" led to Britain publishing the first modern national accounts and GDP in 1941.

By the end of the war it became clear that there would be a need to collect national statistics on behalf of the UN, to support the postwar construction efforts. As a result GDP, now including government expenditure, became the means for governments to manage their economies in the postwar era, including forecasting the effects of their interventions and policy changes.

From 1945 to 1970 world GDP tripled, with the main benefactors being the “rich club” of OECD countries. For New Zealand a buoyant period in the 1950s and 1960s (where average GDP growth was 4.1%) took a sharp downwards turn with the collapse of the wool trade in 1966 and the first oil shock in 1973. New Zealand has since had comparatively modest decades in terms of average GDP growth: 1970s (2.2%), 1980s (2.0%), 1990s (2.2%). The GDP annual growth rate in New Zealand from 1998 until 2018 has averaged 2.61%.

Measuring GDP

GDP remains an abstract construct, which has become increasingly complex as the world economy has developed. GDP can be measured in three ways:

1. Output/production of economy
2. Expenditure of economy
3. Income of the economy

GDP (Gross Domestic Product) which is all the economic output within a country’s boundaries, can be distinguished from GNP (Gross National Product) which measures output and income from overseas. The difference for some small countries can be large. GDP is the sum of all expenditures in the national economy. The calculation is:

• C: is consumer spending (by private individuals or households)
• I: is investment spending (by companies)
• G: is government spending (on goods or services)
• X: is exports
• M: is imports

What GDP Doesn’t Capture

Real GDP per capita is often used as an indicator of standard of living and is useful as it is measured frequently for almost every country in the world. The problems though relate to the weightings which for some countries have not changed in decades – before the full effects of globalization or technological advances were seen. There are also a number of transactions that GDP doesn’t capture: