Data Science: The Modern-day Pillar of Economics
Broad strokes
With the technological advances of recent years, especially since the turn of the millennium, Data Science has become a discipline in its own right, separate from computer science and more closely aligned to statistics. It has carved out a niche for itself where data scientists apply themselves to solving business problems that rely on the access, processing and ultimately the interpretation of data.
This demands a particular skillset such as a good understanding of programming languages, for example Python and R, to help simplify analytics workflows required to access large disparate data sets. The data scientist's skillset combined with that of the economist delivers a winning formula for those looking to distinguish themselves from the herd and master modern economics.
Facts & figures
Findings above are supported by the fact that the prestigious London School of Economics [1] has extended its curriculum in recent years to include an undergraduate degree titled the BSc Data Science and Business Analytics, with the strapline promising learners that they would "learn to analyse data to address real world-problems", real-world problems which are naturally based in economic and business relations.
Another positive indicator is that the former Chief Economist of the World Bank [2] and the joint-winner of the 2018 Nobel Memorial Prize in Economic Sciences, Paul Romer, is a proponent of Jupyter Notebook, an open source web application which allows users to create and share documents including live code, equations and visualisations to support interactive computing across multiple programming languages. This final remark is at the core of Jupyter, the name Jupyter being an acronym meaning Julia, Python and R, all three being programming languages.
For a giant in Economics to be a vocal advocate of a data science tool speaks volumes – no pun intended – and it clearly indicates the direction of travel. As Romer noted on a blogpost back in 2018: "Jupyter rewards transparency; Mathematica rationalizes secrecy. Jupyter encourages individual integrity; Mathematica lets individuals hide behind corporate evasion". [3] Here he is comparing Jupyter with a competing platform, Mathematica, however if we look beyond the descriptives it is quite clear that Romer strongly identifies with the benefits and possibilities which data science brings to the forefront.
Live case study
Shifting to a current real-world problem to illustrate the themes noted above, the UK is approaching its three year anniversary of adopting its post-Brexit EU-UK relationship, working according to its newly imposed trade regulations.
Following the UK's referendum to leave the European Union and by extension the single market on June 23rd 2016, it took over four years as part of a transition period lasting up to 31st December 2021 before the agreed changes were implemented.
Fast forward to the current day, 20th November 2023, other than Brexit serving as a key factor in the success or demise of the UK economy, there are now several other factors which have joined the party which conveniently muddies the waters.
One of the late arrivals to the party is of course the Covid pandemic. Considering this human tragedy purely through a financial lens, Covid led to an unprecedented halt in consumer spending and in turn an increase in the average household's savings, largely brought on in response to office workers transitioning to working from home, all social and leisure arrangements including travel being postponed indefinitely.
Needless to say this contributed to the massive imbalance in finances, at a time when the financial system was already overheated with 0% interest rates thereby limiting options for monetary acrobatics to help ease the economy into a recovery.
That covers two factors, Brexit followed by Covid. As if that wasn't enough, along came the energy crisis triggered by the Ukraine-Russia conflict which started with Russia's invasion of Ukraine on the 24th February 2022. Again, purely considering this humanitarian disaster through an economic lens, the conflict led to the demand for energy outstripping supply and in turn the prices of all impacted consumer goods rocketing.
In summary, these three factors are coming together to create the perfect economic storm which will demand all parties to pull together to sail us through the choppy waters ahead.
With both data science and economics being built on statistics, the key distinction is that data science tends to focus more on predicting future interaction between variables whilst economics tends to focus more on the historic interaction i.e. forward looking vs backward looking, prediction vs causality.
Wearing the economist's hat, I reviewed the trade-to-GDP for the G7 countries leading up to 2023, data and visuals supplied via the Office of National Statistics [4]. It clearly highlights that whilst all G7 members' trade plummeted during the pandemic years of 2020–2022, they all recovered to pre-pandemic levels, bouncing back strongly with exception to the UK which is still lagging behind other members, as illustrated in the visual below:
This illustrates that trade has a reduced weighting as part of the UK's GDP, in relation to its other six G7 counterparts, a valid statement as at Q2 2022. Rather than promoting trade relations outside of the EU as argued by Brexiteers with the run up to the referendum, it appears that the UK's trade openness, in layman's terms its ability to trade, is in fact reducing, insights supported by data science.
Another critical component of a healthy economy is of course the balancing act of inflation, the target being to contain it under 2%. As of October 2023 the UK inflation is set at 4.6%. [5]
If we review data from the 1960s to the current day, specifically tracking inflation it also provides for some interesting insights. The data and visuals below has been sourced from The World Bank group. [6]
The spike we see here in 2008 is where world inflation jumped to 8.9%, in response to the Global Financial Crisis of 2008–2011.
This can also been seen on the equivalent inflation graph for the UK, as illustrated below:
The spike that can be seen in 2008, whilst nowhere as dramatic as what it appears to be on the worldwide inflation scale, jumps to 3.5%, up by 1.1% just the year before and up by 2.1% from the stable level of 1.4 in 2004. [6]
Closing argument
With world and UK inflation levels increasing considerably it supports the notion and news headlines of a recession looming ahead for 2024–2025, if historic trends supported by the data is anything to go by.
Based on the findings above and my personal experience working as a Data Business Analyst within Investment Management, it's evident that data science and economics disciplines are closely interlinked, data science effectively serving as a key pillar of economics and the decisions it drives and supports.
I think it's also fair to conclude that economics is already reaping significant rewards by leveraging new technologies and data interrogation methods available on the market today and everything points to this being a trend which is still in its formative years.
References:
[1] London School of Economics, offering BSc Data Science and Business Analytics degree as part of their curriculum
[2] Paul Romer, Wikipedia
[3] P. Romer, Jupyter, Mathematica, and the Future of the Research Paper (2018)
[4] Office of National Statistics, released 10 October 2022, ONS Website, Recent trends in the international trade flows of G7 economies
[5] Office of National Statistics, released October 2023, ONS Website, Inflation and Price Indices
[6] Inflation data and visuals, World Data Group
