Are wars affordable? The answer of course is ‘yes and no’.

Affording a war is different from financing a war. To make any new thing affordable, either there must be a reallocation of resources or a deployment of resources not otherwise in use. Or a mix of both. Further, resources get destroyed, and not only the resources of the ‘loser’.

Wars may be fully tax-funded – that is, by increased taxation – by one or more belligerents; but most usually they are not. Otherwise, wars are financed. Financing is a mechanism which enables the distribution of spending to differ from the distribution of income. Typically, spending by warring parties exceeds their incomes, so must be financed through government ‘fiscal’ deficits.

Income is the rights to current goods and services; that is, to current output. Present tense. In particular wages, profits, rents, royalties. Finance is the principal mechanism whereby such rights to current output are transferred by some people (including businesses and governments) to other people. By giving up a right to current output, a party either gains a right (ie a ‘claim’) to future output or is fulfilling an obligation – a debt – incurred in the past. Thus, giving up rights to current output is called either ‘saving’ or ‘debt repayment’. Saving, conceding such rights in return for claims on future output, is commonly understood as lending or ‘advancing’ funds.

We note that in many cases, debtors – parties holding obligations incurred in the past – have discretion over when they fulfil their obligations. Likewise, savers (creditors) have some discretion over when they call in (ie realise, spend) their savings; that is, discretion over when they exercise – ie liquidate – their historical claims to current output. As a general matter, is it a good thing if those two matters of debtor and creditor discretion balance out, creating a sense of ‘equilibrium’.

Historically, however, creditors have often failed to liquidate their claims at all; many creditors like to hold onto their claims for indefinite periods, thereby enabling debts to be merely ‘serviced’ rather than repaid. Unrealised claims are called ‘wealth’, and many people like to hold wealth until they die, rather than spend it.

If insufficient current output is purchased by past savers, it becomes a systemic requirement that new debts are contracted and spent.

When sovereign governments contract new debts to fulfil this systemic requirement (possibly as ‘debtors of last resort’), this is ‘fiscal accommodation’. When governments refuse to contract new debts to fulfil this systemic requirement, we may call this either ‘fiscal consolidation’ or ‘public austerity’.

Wars – and preparations for war – may be destructive (or at least non-productive) examples of fiscal accommodation; such accommodating militarisation may achieve that purpose without specific intent to do so. (In the 1930s the renowned economist John Maynard Keynes offered, as an example of contextually beneficial non-productive fiscal accommodation, governments paying workers to dig up holes and fill them in again!)

Wars

Medieval wars were often short-term affairs, because of seasonal patterns of labour demand. Wars have for the most part been labour intensive; and that’s still the case today, even if the casualties of post-modern wars are more likely to be civilians and less likely to be soldiers and sailors.

Medieval sieges often had to be terminated around August because the soldiers in the sieging army had to return to collect the harvest. September was the time of the year when there was virtually zero unemployment. Siege defence was made possible because harvest labour requirements would likely break the stalemate. The corollary is that medieval wars could be afforded because, in late-spring and early-summer, there was seasonally unemployed labour.

In the modern period (approximately 1490 to 1990), especially in Europe, labour became increasingly divorced from agriculture, making it possible to have ever larger standing armies (and navies), making bigger and longer wars possible. Further, the modern period saw the emergence of sovereign nation states; so, increasingly, war finance became intrinsically connected to public finance. Wars of exploitation and territorial expansion became a central feature of the emergent mercantile States.

Public finance and war finance were essentially the same thing in the golden eras of merchant capitalism (roughly 1550 to 1800) and subsequent industrial capitalism. That financial conflation is re-emerging as a new reality of the twentyfirst century, as sovereigns (and their foreign state and non-state proxies) up their military spending while simultaneously diminishing their commitments to the peacetime provision of public goods.

Fast forward to the years from 1989 to 2011

This transition period from modern to post-modern may be seen as a particularly peaceful period – after the Great World War of 1914 to 1945; after the wars of recolonisation and decolonisation which may be seen to have ended in 1979 with the revolution in Iran and Vietnam ending the post-colonial genocide in Cambodia; after the wars in Lebanon, the Falkland Islands, and Iran-Iraq; and after the fall of the empire of the Soviet Union.

The millennial years 1989 to 2011 are sometimes called the ‘unipolar moment’, when the United States could and would call the shots; typically with a foolhardy and exceptionalist perspective of the world as a kind of playpen for Washington and New York largesse. And with a neoliberal outlook through which narratives about the Great Depression and World War Two were recast. In the latter case, World War Two became a grand narration of ‘Hitler versus the Jews’; most of the many other lessons arising from the years 1914 to 1945 were largely forgotten.

I am particularly interested in the affording and financing of the Second Gulf War (essentially 2003 to 2009, an asymmetric war between United States and Iraq); although good starting points are the post-Tiananmen (after 1989) emergence of China and the execution in 1990 by the United States of the First Gulf War.

These charts of financial balances for China and the United States give some important clues about who paid for the Gulf Wars. (For the United States in particular, it is necessary for now, to not be distracted by the dramatic financial accommodations between 2009 and 2021, relating to the Global Financial Crisis and the Covid19 Pandemic.) They show variations over time in private saving and spending, government deficit spending, and these nations’ saver/spender relationships with their outside worlds.

We note that most of the economic and financial cost of war comes after the main event (eg after 1990, and after 2003); as military equipment needs to be replenished, armies need to be expanded, and destruction zones need to be rebuilt. Indeed, the costs of a standing defence force are high whether or not there is a war.

1990, the middle of a period of both economic and financial flux in the world, came at the end of a recovery in the United States following the 1987 sharemarket crash. So, almost unusually, there was no speculative bubble in place, there was increased saving as people looked more to future spending than present spending, and the labour market remained weak.

For the United States, we see in the years from 1991 to 1993, high saving in the private sector – largely household saving – and comparably high spending in the government sector. Thus, domestic private savings directly funded the war. Unemployment in the United States was lower than it otherwise would have been. While savers were not asked whether they were happy that their caution was being translated into government military spending, it’s unlikely that they minded too much; the ‘war against Saddam’ was not an unpopular war in the United States.

In times of recession, when more people than usual are unemployed or underemployed, affording a war is easier but financing a war is harder. Liberal governments must make financial accommodations by departing from the standard fiscal rules they impose upon themselves. (We note that, just this year, 2025, the German Bundestag has made such an accommodation, and abandoned its self-set and dearly-held fiscal rule; giving itself a blank cheque to pursue debt-funded military spending.)

Most modern wars have been afforded through a process of restrained consumption, financed through the mechanism of new government debt and a build-up of household credits; governments owing, and households owning new debt. As a side-effect, and considering the United States, this affording and funding enlarges the combined balance sheet of American banks: more assets (government debts) and more liabilities (private savings).

Affording wars is always a matter of economic resources being deployed into military theatres, whether that is redeployed from civilian production or a reduction of resource underemployment. From a financing point of view, the four options are that wars are funded by taxes (which would not show up on this type of chart), by domestic saving (as happened in the United States from 1991 to 1993), by foreign saving (as happened in the mid-2000s), or by foreign aid from patron to proxy.

War financed by foreign saving may mean direct or indirect foreign funding. Much of the Allies funding in the Great World War was financed by American debt which, in the fullness of time, would be written off; making that war significantly American gift funded, even if at the time the advances were only intended and consented as loans. Nevertheless, the United Kingdom afforded their war only with substantial reductions in normal consumption; this was even more true for most of the other participating nation states.

In the United States chart above, we see (in green) that in every year shown except 1991, the United States has incurred debts to the rest of the world. Though these foreign advances were unusually low in the early 1990s. America’s war in 1990 was domestically funded, and relatively easily afforded.

(We note that that Gulf War involved both an invasion by Iraq and an invasion of Iraq. I make no attempt to discuss the affording or financing of the war from the point of view of either Iraq or Kuwait. Clearly, however, there was a substantial loss and degradation of life in Iraq, and degradation of land and infrastructure.)

The Wars of the 2000s, especially the Second Gulf War from 2003

The United States economy changed dramatically with the birth of the Internet-Age, just after the First Gulf War. Private balances follow a classic ‘bubble’ formation from 1994 to 2000/01; this came to be known as the dotcom bubble, and was characterised by a new ‘information technology’ sector being speculatively debt-financed. Government tax revenues ballooned, leading to unheard-of government budget surpluses. In addition, the United States economy attracted increased foreign credits before the turn of the millennium, though not much then from China.

We can see the collapse of the dotcom bubble in 2001, with a marked reduction in private debt spending, and the ensuing unusually high foreign financing of the United States economy.

The wars of the new-millennium began with the United States’ invasion of Afghanistan in 2001, followed by the bigger United States invasion of Iraq in 2003. These wars were foreign-funded, the US chart shows, and lasted – in their predominant phase – throughout the Bush presidency. (Refer Iraq War troop surge of 2007.)

We can trace this funding of the Bush-wars by examining the China chart. From 2002, we see a clear rise in Chinese private saving and of ‘foreign investment’. The ‘rest of the world’ percentages represent spending in the rest of the world (from China’s perspective) made possible by non-spending in China.

At its peak, China’s foreign investment ‘current account surplus’ – for our purpose, China’s excess of exports over imports – reached almost 10% of GDP in 2007. This co-dependency of Chinese exports and American imports has been called by some Chimerica; the best known proponent of this concept is the British global historian Niall Ferguson.

As well as considering the percentages, we remind ourselves that Chinese supercharged annual economic growth, which bottomed-out at 8% in 1999-2001, climbed to 14% in 2007. Given earlier growth in the 1980s and 1990s, China was no longer starting from a low base. These were massively increased levels of Chinese economic output in the 2000s; output sent from China rather than spent in China.

The result was that industrial capacity within the United States was freed up to supply military goods rather than civilian goods. While China provided the ‘butter’ (ie consumer goods), Uncle Sam was freed to specialise in the production and deployment of ‘guns’.

While China paid for the Second Gulf War through its massive export surpluses, for the West in general and the United States in particular, the war was fought for free; a ‘free lunch’ so to speak. Of course it wasn’t technically free; China built up a massive amount of financial claims on the United States, though it was never clear how or when China might exercise those claims. China is yet to show any desire to acquire the American imports which would constitute the settlement of China’s claims on the United States.

China will only be reimbursed for its massive lending to the United States in 2003 to 2008 when we see, in its future financial balances’ chart, a whole lot of green on the upper ‘savers’ side. Otherwise, China’s loans to the United States will morph into gifts. An export surplus can only be reimbursed in the form of an export deficit; not China’s style in current or near-future times.

Tax Cuts

Not only did the United States wage two major wars in West Asia, close to America’s Indian Ocean antipodes, it did the unheard-of for a nation at war; it reduced its tax rates. While the most obvious way to fund a war is to raise taxes, the United States did the precise opposite; to not fund the wars ‘because it could’. China was happily paying for those American wars. For many Americans not directly involved, these wars were more than a ‘free lunch’; they were, through tax cuts, a ‘sugar hit’.

Indeed, this detachment of fighting from cost-bearing has become the most dangerous facet of the emergent ‘Warrior epoch’. Western elites have come to believe that they can undertake wars – be they ‘good wars’ or ‘bad wars’ – without themselves facing up to the reality that all wars are costly.

The United States legislated two major rounds of tax cuts, in 2001 and 2003. The first round was undertaken in the light of the Clinton budget surpluses (see the year 2000), and without awareness that war was coming. Those Clinton fiscal surpluses were unsustainable, a consequence of the dotcom bubble mini-boom, though the tax cuts (ill-targeted as they were) helped to fiscally accommodate the recovery from the 2000/01 dotcom bust.

The 2003 federal tax cuts were inexcusable. Initiated just as the pre-Gulf-War hype was peaking, these tax cuts passed through Congress and the Senate during the peak initial phases of the war. The incongruence of simultaneous military aggression at scale and tax decreases was astounding in its brazenness.

After 2011

The principal wars in the 2010s were located in Afghanistan and Syria; there was additional militarisation in Eastern Europe and the former Soviet Union, associated with the eastward expansion of Nato.

China played a constructive new role in that decade.

An important feature of global financial imbalances – very clear in the American chart – was the Global Financial Crisis, showing resurgent American private saving (mainly debt repayment) and the spectacular (and necessary) US Government accommodation of that dramatic change in private behaviour. Then we see a return to normality from 2013 to 2019. Higher than usual United States government deficits were a critical part of the global recovery from the financial crisis. (We may mention in passing that the New Zealand government’s fiscal policy – under National and Labour – has been and still is non-accommodating; the pandemic year 2020 being the exception that ‘proves’ the rule.)

For the second critical component of the 2010s’ global economic recovery, we can see a big change in China’s financial balances. In particular, we see the emergence of the Chinese consumer and taxpayer (much less blue and less red). And Chinese net exports substantially diminished as a share of the Chinese economy. Consumer spending and government spending in China and the other BRICsrevived global demand for non-military goods and services.

Although the United States incurred a specific debt to China during the Second Gulf War in the 2000s, subsequently the whole West ‘owes’ China a considerable debt of gratitude for its role in restarting the global economy around 2010. Thankyous to China have been considerably lacking, however, as the West increasingly seeks to point its military hardware at China.

The West – led by the United States – has gamified war, and has become indifferent to non-western lives. There are also too many signs that western elites are becoming indifferent to western working-class lives; starting with indifference to the many immigrants who are already performing so much of the necessary labour to support higher-middle-class living standards.

China, already on the verge of a balance-sheet recession in the view of Richard Koo, may now be following in the financial and economic footsteps of Japan in the 1990s (see my Red Gold; Japan’s lesson for the world). Certainly China’s financial balances’ chart (above) is starting to look very Japanese, with a smallish and stable export surplus, and large government deficits.

2020s’ Wars

After the 2020 Covid19 Financial Crisis, which, as in 2009, required huge fiscal accommodations – especially by the United States federal government – wars have become proxy affairs whereby the means of war have been largely gifted by patrons to their proxies. Such financing leaves only small marks on countries’ financial balances charts. Though the patron nations will have larger-than-otherwise government deficits; see the United States’ government balance (above) for 2023 and 2025 (and the 2025 forecast).

The financing of the two sides of the Sudan ‘Civil War’ appears too convoluted to examine here. It would seem to involve proxies of proxies, and to be an important outlet for internationally traded military goods.

For the West, the affording of the wars in Ukraine and Israel-Palestine would appear to be mainly through a mix of gifts and loans by patron governments, meaning involved governments undersupplying too few peacetime public goods. (Too little ‘butter’, to use that metaphor, and too many guns.)

Russian citizens will be incurring substantial opportunity costs, mainly through higher taxes, a reallocation of government spending, and reduced opportunities for its citizens to live international lives. Ukraine seems to be funding its war through a mix of foreign gifting and government debt; though its people – like Russians – have been paying a high price through reductions in their living standards.

Israel continues to be a net exporter, so its deliveries of military hardware from the United States should definitely be regarded as aid rather than imports. Lucky Israel! To be able to fight its neighbours on such favourable terms is a privilege rarely granted.

In Retrospect

Wars are costly. Very intensive and extensive in the use of resources and the destruction of resources; let alone the loss of quantity and quality of life.

In all wars, all parties incur costs; significant costs. Sometimes, a party to a war can avoid most of those costs through having someone else pay. Of course, the United States paid to some extent for the wars against Iraq in terms of American lives lost and degraded; little cost was borne by those Americans who propagated those wars, though.

The material costs of the wars in the 2000s were paid – indirectly – by Chinese households not consuming large swathes of the goods they produced; Chinese workers and capitalists were, on an increasingly massive scale, exporting the fruits of their labour and their capital to the United States. More sending than spending. Much more. (A Marxian analysis would attribute the seemingly costless affording of the US-Iraq war to the extraction of ‘surplus value’ from the Chinese working class by the American capitalist class.)

Yet these Chinese costpayers didn’t much mind, because – while their abilities to enjoy the increasing fruits of their labours were highly constrained by China’s export policy – they were happily stacking up claims on future production; deferred enjoyment, rather than the pure exploitation which occurred in the early years of Chinese Communism.

China bore the West’s costs in other ways too; in those years Chinese people suffered huge environmental costs, at a time when natural environments were improving in the deindustrialising West.

There was a wider set of ongoing costs, however, arising from the ensuing highly unbalanced global capitalism. United States’ industrial survival is now largely dependent on its specialisation in military hardware and software; meaning that the United States’ economic deformation has made that country into a predatory warrior state. Violences, especially upon non-Americans, are today directly committed by the American state; and through both exported and gifted military goods and services, and through violations committed directly by America’s proxies (and, as in Sudan, by its proxies’ proxies).

Wars, when they happen, are affordable because they happened. They are very costly, both in terms of their opportunity costs (the loss of other uses to which the deployed resources could have been put) and the human misery of death, destruction of habitat and taonga, and injury. They are commonly financed by third parties – eg Chinese households – who may or may not enjoy reimbursement for their credit advanced.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The chart above summarises Japan’s financial balance sheet since 1980. A wall of red below the line, and blue above. Additionally, a persistent ‘slice’ of green below the line, indicating that Japan – the country, not the government – is very much a creditor (ie saver) nation.

This red wall has been the norm for Japan, except for a brief period in the late 1980s and early 1990s when Japan had one of the world’s most spectacular financial bubbles and busts. Japan took a decade to get over that crisis, and in the process forged a new macroeconomics; a macroeconomics created ‘on the fly’ so to speak, and which substantially demonstrates the validity of modern monetary theory. (For Japan’s story of recovery, restoration and education, refer The Holy Grail of Macroeconomics: Lessons from Japan’s Great Recession 2009 by Richard Koo.)

The red wall shown in the chart is Japan’s monetary base. It functions in the monetary world much as gold was meant to function during the gold standard era. Japan is not starved of money, and Japan has inflation no higher than the rest of the most-economically-developed world. Its equivalent of New Zealand’s Official Cash Rate is 0.5%; the highest it’s been for over ten years. (Contrast the OCR, which has come down from 5.5% to 3.0% in the last year-and-a-bit.)

Further Japan’s ratio of people over 60 to people under 60 reached ‘crisis levels’ at least a decade ago, yet Japan is still able to provide for – to afford – an older generation of healthy and happy retired people.

The wall of ‘red gold’ in the chart features in Japan’s national accounts as government debt. On the other side of Japan’s national balance sheet, that red wall becomes a blue wall; a blue wall which features as the principal store of private wealth in Japan. The red wall and the blue wall are the same wall; it’s simply ‘painted’ blue on one side and red on the other.

During 25 years of construction, that red wall grew to 250 percent of Japan’s GDP in 2020; to the equivalent of five trillion United States dollars. Since 2020, Japan’s government has continued to run substantial though diminishing deficits, adding to its red golden wall of public debt; though the growth in the wall since 2020 has been slower than the growth of Japan’s economy, meaning that Japan’s government debt has fallen to 237 percent of GDP.

(Among developed economies, after Japan the next biggest red walls are those of Singapore, Greece, Italy, and the United States. The Singaporean government effectively borrows internationally at very low interest rates (and in its own currency), thereby providing much of the funding for its much-vaunted sovereign wealth fund; and the United States’ wall of red gold is the de facto base of the global monetary system. The currencies of Singapore and Japan will be best placed to take over from the $US as world reserve currencies – because of, not despite, the level of public debt in those currencies – should any financial catastrophe befall the USA.)

Japan’s red gold is safer than yellow gold, because it is fully backed by the Japanese taxpayer; all governments have the sovereign right to claim taxes from their citizens. The purchasing power of Japan’s store of red gold is not contingent on the variability of the global market for yellow gold.

The Yellow-Gold Bug

Meanwhile, the world has become enamoured with yellow gold; you only have to look at the triffid-like growth of yellow bling in the White House in Washington DC. Likely the unsanctioned new ballroom replacing the East Wing of that presidential residence will soon enough house the world’s biggest display of yellow gold. Red gold drives the global economy forward, though – given the levels of inequality outside of Japan – into some nasty elite consumption spaces. Yellow gold drives the narcissists’ vanity, and is driven by the widespread fear of the middle classes – especially in the emerging and developing world – of global collapse.

Also meanwhile – just this week – Sudan has experienced another genocidal setback; see Massacre in el-Fasher: What’s happening in Sudan right now? and ‘A true genocide’: RSF kills ‘at least 1,500 people’ in Sudan’s el-Fasher, both Al Jazeera 29 October 2025. Many people will be wondering what it’s all about, though listening to mainstream news media won’t help very much. (We know that Sudan will continue to play a role in the world’s military armaments industry, both as a cynical testing ground for big boys’ toys and as a significant future supplier of the rare earth minerals needed to make those toys.) However, this piece from Al Jazeera’s Charles Stratford gets close to the real truth: Sudan’s natural wealth becomes the new front line of its ongoing war, YouTube, 29 October 2025.

Sudan has produced truly massive amounts of gold bullion since the start of its ‘civil’ war; yellow gold which has been smuggled out by the genocidal ‘Rapid Support Forces’ under the patronage and tutelage of its megarich backers. ‘A true genocide’: RSF kills ‘at least 1,500 people’ in Sudan’s el-Fasher states: “Saudi Arabia, Egypt, Qatar, Türkiye and Jordan have condemned the abuses committed by the RSF in Sudan”. Significantly missing from that list – and from the whole story that we do hear – is the UAE, the United Arab Emirates, reputed to be the RSF’s principal patron.

The Virtual-Gold Bug

Crypto-currencies seek to mimic gold, through an equally environmentally unsustainable process of ‘mining’. They have succeeded, becoming a speculative commodity par excellence. Indeed the First Family, in addition to its very overt and rather sickening displays of yellow gold, is reputed to have made nearly a billion dollar’s worth (Reuters) of acquisitions (and capital gains) of various cryptocurrency hoards. (And see this from the Australian ABC.)

New Zealand’s Opposition leader Chris Hipkins wants to make revenue from a capital gains tax on residential and commercial real estate (see this on Scoop). Yet, except for a brief bubble in 2021/22, there have been minimal capital gains on New Zealand land holdings since 2017; rather, capital losses have been the 2020s’ norm. Yet there are massive capital gains being made, in yellow and especially in virtual gold. Also, there are increasing claims that world sharemarkets are at unsustainable levels; see Awash with cash. How the investment world is feeding upon itself, ABC. Indeed, the ‘investor’-class is busier than ever, though not in real estate. Further, there is no clear reason why there should be a resumption of real estate bubbles anytime soon, given the abundance of alternatives.

These booms in real-gold and unreal-gold booms pose a major financial instability risk. Red-gold, on the other hand, can be the epitome of stability.

Conclusion

There is a form of ‘gold’ – invisible in the political chatter – which represents the backbone of the world’s monetary system. That’s red gold, and Japan is showing us the way, if we could only look and see. (There’s an ever-present fear that Japan will sooner or later snatch monetary defeat from the jaws of victory. And see this about “embarrassing antics in Tokyo“: I hope that Japan’s new prime minister Sanae Takaichi will not push too far her Mrs Thatcher reputation.)

Meanwhile, parts of the financial world are spinning out into some Lulu La La land, in the gargantuanly wasteful – and at times genocidal – pursuit of gold and virtual gold.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Earlier this week, in The Truth about Prices in New Zealand, (on Evening Report, and on Scoop), I showed how the Consumers Price Index (CPI) is a lagging measure of inflation, and that the Producers Price Index (and the use of six-monthly rather than annual data) gives more timely information about turning points in inflation.

In this chart I look at the two published components of the CPI: inflation in the “tradable” and “non-tradable” sectors. And, because the CPI is a lagging measure, I have used the best presentation for historical reflection rather than for timely headlines. (This measure is annual, and it compares whole 12-month periods with each previous whole 12-month period. The latest data point in the chart matches to 31 March 2025, which is at the centre of the year-ending 30 September 2025.)

Consumer prices are retail prices, whereas producer prices are wholesale prices. Tradable CPI inflation mainly represents the retail prices of traded goods; goods New Zealand mainly exports and goods New Zealand mainly imports. Much of their pricing includes the markups of domestic retail outlets, domestic transport, and domestic customisation and packaging of imported manufactures and foodstuffs.

The non-tradable sector is mainly domestic services, utilities, and construction. Sellers of these services and goods do not compete with foreign suppliers.

In addition, the chart shows the monetary policy setting of the Official Cash [Interest] Rate; set by the Reserve Bank.

Interpretation

In the 2000s the OCR was set at what was then understood as normal levels, in the order of six percent. The received narrative of the time was that interest rates should be significantly above inflation rates. The achievement of such high real interest rates, then (and globally, not just in New Zealand), was probably the main single cause of the subsequent Global Financial Crisis in 2008. It created an environment in which money was transmitted en masse from borrowers to savers, and the ‘investor-class’ enjoyed ever-increasing demand for financial assets which would supercharge their ‘paper’ returns. In those years there was nothing like the degree of debt-phobia that exists today; leverage was the name of the game.

Despite the dubious anti-inflationary narrative which justified these high interest rate settings, high interest rates did not force countries’ domestic inflation rates towards the target rate of two percent. Due to globalised competition, the wholesale price inflation of traded goods remained low; even negative at times. In addition, for individual countries such as New Zealand, exchange rate appreciation also served to keep ‘tradable inflation’ very low. Indeed it was those high interest rates which facilitated the currency appreciations of ‘commodity currencies’ such as the New Zealand dollar.

Nevertheless, by the late 2000s, high ‘wealth effects’ – including (indeed especially) among indebted home owners – saw world commodity prices soar, leading to high inflation in the tradable sector despite commodity currency appreciations.

The Global Financial Crisis of 2008 and 2009 saw the collapse of a number of financial asset prices worldwide. Central banks cut interest rates dramatically, to revive a failing and flailing world economy. New Zealand’s annual inflation fell to below two percent. And it stayed below two percent for more than a decade. While non-tradable inflation sat between two and three percent, tradable inflation brought the overall CPI to one percent and even less (especially in 2015, ten years ago). (Note that the 2011 inflation ‘spike’ in New Zealand was due to the increase in the rate of Goods and Services Tax; GST.)

Interest rates in New Zealand were slashed to one percent in 2019, in the belief that that change would induce an upwards movement in the rate of inflation; and in the full knowledge that – if such monetary policy changes were to be effective – there would be a time lag of at least one year between the policy and the outcome.

Inflation did increase in 2021, though it would be foolish to attribute much of that to reductions in the OCR. Reductions on the OCR from 2015 had had minimal if any impact on inflation. And we know that 2021 and 2022 were very trying times indeed in the world’s supply chains, with pandemic and war. The supply chains quickly adjusted however.

Tradable inflation – even for a lagging measure such as the CPI – clearly turned downwards in 2022. And was plummeting in 2023. The steep rise in interest rates in 2022 could not have been the cause of the substantial tradable disinflation in retail prices; a falling inflation which began about the same time as the monetary policy squeeze.

Those increases in the Official Cash Rate almost certainly had an impact on non-tradable inflation, however. But not in a good way! Just as OCR settings between five and six percent in the early 2000s seem to have held non-tradable inflation at high above-target levels, so also do they seem to have held non-tradable prices in New Zealand in 2024 (and, based on recent quarterly data, continuing to have such an impact in 2025) at distress-inducing above-target levels.

The cost-of-living crisis since the National-led government took office is both the direct effect of the counterproductively high interest rates, and the laggard high non-tradables CPI-inflation which has extended well into 2025.

The real agenda for high interest rates would appear to be to recreate the early 2000s’ financial environment, whereby interest rates were well above inflation, and the elites of New Zealand and elsewhere were embarking on their ultimately destructive journey of inflating paper-wealth.

______________

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The first chart shows annual price increases in New Zealand for businesses (PPI: Producers Price Index) and consumers (CPI: Consumers Price Index), since 1999. We note that the latest CPI datapoint is for the third quarter of 2025, meaning that it’s centred on mid-August. The most recent PPI data is for the second quarter of 2025, meaning that it’s centred on mid-May.

For the whole period, the first important points to note are that the PPI is more sensitive to changing influences on prices than the CPI, and that the PPI tends to lead the CPI. Indeed, annual CPI inflation is a lagging measure of price change; meaning that it’s a poor measure to base policy decisions on.

The other key point to note is the unusually long lag of the CPI after 2022. Using the more sensitive and timely PPI measure of price inflation, we see that inflation in New Zealand troughed in 2023, and that, using the ‘outputs’ PPI, annual inflation in New Zealand was bang-on the two percent policy target at the time of the 2023 general election.

Despite the claims of our Prime Minister that he inherited “seven percent inflation” from the previous Labour government, in the two years since the election, actual inflation (based on the more sensitive PPI) has been rising.

It is very clear that there was a double ‘price spike’ in 2021 and 2022, periods exactly corresponding to the disruptions to global supply chains caused first by the Covid19 pandemic and secondly by the Russia-Ukraine war. Commodity price increases (PPI-inputs) fell almost to one-percent once those global supply disruptions were resolved. After that, the main source of ‘cost-of-living’ increases – suggested by the CPI lag in 2024 – was panicked and counterproductive domestic policy measures.

Historically, we note that, at the onset of the 2008 Global Financial Crisis, inflation in New Zealand was far worse than anyone realised at the time. We also note that, while the 2011 ‘spike’ in CPI-inflation was due mainly to the increase in the GST rate, there was also a spike in producer price inflation at that time. Normally the amplitude of PPI-inflation is greater than for CPI-inflation; because of GST, this amplitude difference did not happen.

Best leading measure of price variation: biannual price change

The most-timely measure of price variation is the quarterly change of the PPI [inputs]. However, quarterly measures are notoriously ‘noisy’, so the first reliable measure of price variation is the six-monthly [ie biannual] change in prices. The measure here takes six months (the two most recent quarters, averaged) compared to the previous six months.

The second and third charts clearly show both the annual and annualised biannual rates of PPI-inflation. The chart clearly shows how the six-monthly (biannual) inflation rate reveals the key inflation turning points first. By the August 2023 release of the PPI data, it was evident that – by the first available measure of prices – inflation in the first half of 2023 was below two percent. Yet, in election year, the Labour Government never mentioned this very favourable piece of economic news! Why was the actual data not being discussed? Presumably because the truth conflicted with the narrative about inflation; the narrative which New Zealand society succumbed to and was cowered by. Part of the problem is the time-poor (and sometimes credulous) media having been, in effect, trained to follow certain statistical indicators but not others.

These charts also plot the Official Cash Rate (OCR), the principal (though typically misplaced) policy lever to push-back on inflation and deflation. They show that anti-inflation policy commenced late in 2021, and peaked in 2023 and 2024. Thus, the ‘anti-inflation’ policy was persevered with well-after the leading indicators had shown that the inflation problem had disappeared.

The fourth chart shows the annual and annualised biannual rates of increase of consumer prices, again showing the OCR as well. Once again, even though the CPI is a lagging price-level indicator, a proper look at the CPI data shows that CPI-inflation was falling markedly in 2023, and that there was no case for anti-inflation policy in late 2023.

The explanation for the unusually long lag of the CPI (compared to the PPI) lies in the fact that the perseverance of anti-inflation policy itself created an ongoing ‘cost-of-living crisis’. If we go back to the first chart shown, it is the long lag in CPI inflation in late 2023 and in 2024 that is in fact the essence of the ‘cost-of-living crisis’. Rather than the crisis being cured by the contractionary monetary policy settings (of the OCR), that extended CPI lag was caused by the anti-inflation policy.

The fifth chart goes back to the PPI-inflation, using the ‘outputs’ measure of business prices, as in the third chart. This chart shows the OCR settings shifted by 18-months, to simulate an 18-month lag. The reason for this is that we are told that monetary policy takes at least a year – with Reserve Bank research in the late 1980s claiming as much as five years – before monetary policy ‘does its work’. If 18 months is the correct lag between policy and outcome, then we should see upturns in the OCR coinciding with downturns in inflation; and downturns in the OCR coinciding with upturns in inflation.

Instead, considering the two years from mid-2023, we see the very opposite, the upturn of the OCR almost exactly coinciding with the upturn of business inflation. We know that the short inflation spike of 2021 and 2022 was caused by global supply-chain disruptions; this kind of causation is probably true of some other inflation spikes. Also exchange rate fluctuations contribute to spikes in price variation. If we look at the late 2010s, we see falling interest rates accompanying falling rates of price increase; contra to the policy narrative. In the early 2010s we see fluctuating inflation while interest rates were essentially unchanging. In the late 2000s, we see interest rate increases matching inflation increases; again, contra to the policymakers’ narrative.

The conventional neoliberal narrative about inflation is that there is a substantial lag in policy effectiveness, and that inflation is principally driven by expectations of inflation. In this narrative, the inflation data should not be ‘spiky’ at all; rather, once set in, inflation supposedly establishes its own momentum or inertia. The PPI data clearly refutes this ‘momentum’ narrative; inflation is not driven by expectations arising from immediate past inflation. And the alleged momentum in CPI-inflation in New Zealand in 2024 is clearly false; rather there was a lag in CPI-disinflation caused by interest rates being too high; not too low. (Disinflation is falling inflation, whereas deflation is falling prices.)

Accepted reasons for an OCR-increase to PPI-outcome lag include the fact that business loans – like home loans – are typically set at fixed interest rates, for say two or three years. In the case of a falling OCR, however, businesses may quickly repay (or break) high interest loans and refinance as quickly as possible with the new low interest rates. So, policy reductions in the OCR are likely to affect outcomes more quickly than increases in the OCR.

It is looking as if anti-inflation policy actually achieves a mix of neutral and pro-inflationary outcomes. My suspicion is that anti-inflation policy is substantially pro-inflationary – counterproductive – with a lag of 15-24 months; and anti-deflation policy is actually pro-deflationary, with a shorter lag.

It is likely that the lag from anti-deflation policy (as in these years: 2001, 2008, 2015, 2019; refer third chart) to consequential deflation is quite short, in large part because the commencement of disinflation commonly precedes the policy. (Like trying to end a war that’s already ending.)

2026 and 2027 will be interesting because the longer outcome lag from high interest rates in 2024 and the shorter outcome lag from falling interest rates in 2025 suggests a wait until later in 2026 before there are marked falls in PPI-inflation, and early 2027 before marked falls in CPI-inflation.

There are at least two disruptors, however, given the global environment in flux. First is the 2025 American-led haphazard disruption to the already disrupted global economy, including the redirection of global supply-chains in favour of military goods and services. Second, for New Zealand, there is the ever-present possibility of a domestic financial crisis which would see a rapid fall in the value of the New Zealand dollar and therefore a 2027 spike in high rather than low inflation.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

An increasing proportion of New Zealand’s immigrants are foreign citizens. In the 2010s – especially the later 2010s – a critical driver of immigration had been returning New Zealand citizens. As the headlines have indicated, that process of sourcing immigrants from the New Zealand diaspora has long finished.

Where have New Zealand’s post-covid immigrants come from? The following table shows immigration from the 31 countries which Statistics New Zealand follows. The estimates for the years-ended-August have just been released.

We note that not all intended migrations to New Zealand are successful. Most immigrants arrive on non-residence visas, and then have to apply for permanent residence or other long-stay visas. Unsuccessful immigrations arise both from failures to secure the desired permission, or from immigrants themselves having second thoughts. There are two possible outcomes of unsuccessful immigration: return migration, or onward migration.

Onward migration may take place following immigrants’ success in gaining New Zealand passports. But that is not unsuccessful immigration, and it’s not shown here. The data below looks at the 12-month period ending August 2023, and deducts the migrant departures for each nationality in the following 12 months (ending August 2024). For comparison, the table also shows 12-month period ending August 2024, deducting the migrant departures for each nationality in the 12 months ending August 2025.

These data are estimates for successful immigration (as defined above) by migrants’ nationalities:

It turns out that Philippines is the 2023 ‘winner’. Philippines consistently has few return or onward migrants. We note that the Philippines’ number dropped more in 2024 compared to India, probably reflecting the larger numbers of Indian migrants who arrived as tertiary students.

Two other stand-out immigrant countries – relative to their source populations – are Sri Lanka and Nepal.

The dominant groups of countries are our Pacific neighbours (Oceania); and South and East Asia. In this context we should note that a substantial majority of immigrants from Philippines, Malaysia, and Indonesia are ethnic ‘Austronesians’, the same broad ethnic group as our indigenous Māori and most of our Oceanian immigrants. Immigrants from Philippines are a particularly good fit, because of their similar Christian culture and because they are ethnic cousins of indigenous Aotearoans.

That’s not to say that any other national group is a bad fit. Most of our immigrants seek to integrate sufficiently to become Kiwis, without being under pressure to assimilate into Euro-Kiwi norms. Interestingly, of the six top immigrant-source countries, New Zealand only has direct flights with two: China and Fiji.

We note that the richer Asian nations feature well down the list. And we note the disproportionately low representation of nationalities with mainly Muslim populations. Indonesia, with 2½ times the population of Philippines has only 2½ percent of the Philippines’ successful immigration. Indonesia, our near-invisible near-neighbour, is the fourth most populous country in the world, and may well have more people than the United States by 2050.

With slightly more immigrants than from Indonesia is Pakistan, the world’s fifth most populous country, and a country with strong sporting links to New Zealand. But Pakistan is way below India in the above table. A surprising omission from the table is Bangladesh, the world’s eighth most populous country, with more residents than Russia (the world’s number nine). Bangladesh does have a significant community in New Zealand, including my GP doctor. I suspect that Bangladeshis feature strongly in the ‘other Asia’ category, along with Cambodians who continue to operate small bakeries in Aotearoa New Zealand. Another country of importance missing from the list is Singapore, whose airline does bring many if not most of our South Asian immigrants.

Other countries not mentioned so far in the world’s top-ten by population are Brazil, Nigeria, and Mexico. Of these only Brazil features in the table above, although Nigeria may well have a significant presence in ‘other’, and Mexico has had some high-profile immigrants to Aotearoa New Zealand. Brazilian immigration, which appears to be dropping off, may return once China Eastern commences flights from Auckland to Buenos Aires.

We see the richer countries in Europe and the Americas (traditional sources of immigration), and Australia, feature in the bottom half of the ‘Top-31′; much more so for 2024 than for 2023. We note that the negative numbers in 2024 mean that more people with those countries’ passports departed in 2025 than arrived in 2024.

Ukraine doesn’t feature, though it might be a major part of ‘other Europe’. Czechia, which I am surprised Stats NZ have included, may be taken as a proxy for Eastern Europe. Also, ‘other Africa’ has held up while South African successful immigration has halved.

The data all reinforces the fact that New Zealand is a demographic turnover country, with the momentum of immigration coming from much poorer non-Muslim countries, and with a significant outflow of richer-country migrants.

For some up-to-date perspective, the table below shows estimated immigration for the featured countries in the year to August 2025. It shows an increase in migrant arrivals from some richer countries, such as United States, Australia, Japan, Germany and France; however, it is likely that similar numbers of these nationalities will leave New Zealand in the next 12 months as arrived in the previous 12 months. Many from France will actually be from New Caledonia; from Oceania rather than from Europe.

Finally, total arrivals of foreigner immigrants were 201,950 in the year to August 2023; 142,661 in the year to August 2024; and 112,591 in the year to August 2025; much lower than immediately post-covid, but still high. Total departures of foreigner immigrants were 35,972 in the year to August 2023; 46,099 in the year to August 2024; and 54,092 in the year to August 2025.

So, in the last year, foreigner migrant departures from New Zealand had reached almost half of foreigner migrant arrivals. This suggests that, for many, immigration to New Zealand is a fraught and often unsuccessful experience.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The following tables represent an update of mortality by sex in relation to Table 2 from Decennial Increases in Deaths by Birth Cohort, in Aotearoa New Zealand.

By looking at deaths registered in February to May only, it is possible to extend trends into 2025, avoiding fluctuations arising from winter illnesses.

The numbers look at people born over a ten-year period and the percentage increase in deaths in a given recent year compared to ten-years earlier. I am most interested in the ‘generations’ born between 1935 and 1990. The oldest generation/cohort shown will not have many more deaths than ten years earlier, because more than half have already died before the age of ninety. For younger generations, only a small minority have already died, meaning that a population can be readily compared with its younger self.

Results are unreliable for people under 25, because too few of them die to reveal any patterns.

Typically, at least for working-age adults – defining working age here to mean about 25 to about 75 – a birth cohort will normally have about 100% more deaths in a given year (eg 2020) compared to ten years previously. We can see that in the Male table below, by looking at the 2010-2020 column, and by looking at the 1935-45+ row.

If we look at the first highlighted figure of 98.3%, it means that 98.3% more men born from 1945 to 1954 died in 2020 than in 2010. The next figure in that row says that 100.2% more men born from 1946 to 1955 died in 2021 than in 2011. The last figure for that row says that 119.4% more men born from 1950 to 1959 died in 2025 than in 2015. (Noting again, that these data are for February to May only.)

This decade we have observed some problematic increases in deaths for men born between 1955 and 1980. (I would rate any number over 120% as ‘problematic’.)

For Females, these increases in death numbers over ten years for a generation/cohort are even more concerning; though it remains true that fewer working-age females are dying than working-age males. It’s more that women are catching up to men. As with men, it is those women born between 1955 and 1980 where the greatest concern lies.

While I’ll leave these data for others to interpret further, the numbers tend to bely the mantra we hear from the finance industry and many politicians that “we are all living longer”. The aging process seems to be coming earlier for people born after 1955 than for people born before that year. (It’s too early to say whether this conclusion about ‘Gen-X’ will also apply to ‘Gen-Y’. While Gen-Y men, born after 1975, seem to be doing OK so far, data for Gen-Y females is not looking too good.)

A final point to note is that Aotearoa’s working-age population is particularly affected by immigration and emigration. The numbers given here will be distorted if, for any cohort within that ten-year period of comparison, there was a marked difference in emigration compared to immigration. We should note, however, that both immigrants and emigrants (to and from Aotearoa New Zealand) tend to be healthier than average for their birth cohorts. Thus, data of this type – which does not rely on population denominators – can reveal subtle truths which may otherwise remain hidden.

Ultimately, societal problems – such as inequality, insecure housing, over- and under-work – all do have an impact on average lifespan at least as much (if not more) than the state of a country’s healthcare services. Actuarial methods of measuring lifespan are lagging indicators of the health of a national population. And they can be problematic in this country, because so many people who will die in New Zealand were not born here; and vice versa.

My tables are forward-looking rather than backward-looking. They warn of trouble ahead, especially in relation to those people born in the 1970s and 1980s. We may note the following: What researchers suspect may be fuelling cancer among millennials (Washington Post article, published in NZ Herald on 30 September 2025).

*******

Data is from Statistics New Zealand, Births and deaths: Year ended June 2025. That data series only begins in 2010.

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

In the wake of the recent release of labour force data (Household Labour Force Survey, HLFS, Nicola Willis bemoans ‘glass half empty’ view of unemployment figures, RNZ 6 August 2025), 1918-1920 National Party Leader Simon Bridges, has called for economic “stimulus” to rescue in particular the dire Auckland economy. (See Call for government to help Auckland as unemployment rises, RNZ; contrast the Minister of Finance Nicola Willis’s retrospective and ongoing advocation of fiscal suffocation Dangers of Excessive Spending Highlighted, Scoop; both 7 August 2025.)

My focus here is to look at the historical and recent employment rates of older women (aged over 55), and to consider the importance of their spending to the health or otherwise of the New Zealand economy. My reference is the first chart highlighted in Employment in New Zealand – especially of women – at the Age Margins, Evening Report, 7 August 2025.

The chart shows that there is a huge increase in the percentage of older women who meet the official definition of employment. (This generous definition includes wage/salary workers – fulltime or part-time – self-employed workers, active employers, and people working without wages in a family business.) The data reveals a huge increase in the ‘participation rate’ of older women in the labour market.

The age group 60-64 had a particular impetus to retire later, namely the rise in the early 1990s of the age of entitlement to New Zealand Superannuation from age 60 to age 65. But the pattern is essentially the same also for women in their late fifties and in their late sixties.

The appropriate benchmark year is 1987, by time the HLFS was bedded in and before the economic consequences of the financial crash in late 1987. While the high period for employment of older women is 2022 or 2023, when jobs were plentiful, we can be sure that the actual participation rate has not fallen since 2022, and has probably continued to rise. (We can disregard participation rates published in the HLFS; they are based on definitions of unemployment which only realistically apply to men aged 30 to 60. There is much ‘hidden unemployment’ amongst older women.)

For women aged 55-59, we see a rise in labour market activity from 43 percent to 80% in 2018 and 2023. For women aged 60-64, we see a rise in labour market activity from 18 percent to 70% in 2022. (The dip for this early-sixties age group in the late 1980s and early 1990s is unemployment masquerading as ‘retirement’.)

For women aged 65-69, we see a rise in labour market activity from 8 percent to 44% in 2022. For women aged over 70, we see a tenfold rise in labour market activity from 1995 to 2025. (We desperately need a ’70-74′ age category in the published data; this ‘early-seventies’ cohort is likely to now be New Zealand’s fastest growing employment demographic.)

Overall, this truly massive labour force participation of older women in the last thirty years has been a barely noticed social revolution. The increase of employed older women is even more dramatic than these figures look, because New Zealand’s highest birth numbers were in the late 1950s and the early 1960s. These women are now in their sixties, and born with higher life-expectancies than their parents.

It seems unlikely that this increased labour force participation is a result of the rise of feminism in the 1970s; an increased advocacy for paid work was one plank of that feminism. Though feminism may have played a significant but lesser role in this huge social change. It seems far more likely that the main driving force is economic pressure upon households; stresses that have increasingly required all adult household members to be attached to the labour force, rather than the pre-1980s’ emphasis on an individual (typically male) ‘breadwinner’.

The stresses initially hit households hardest in the late 1980s through massive rises in mortgage interest rates, and in the more frequent revision of interest rates by banks during the lifespans of home loans. To that we can add an increased reliance on other forms of personal debt, such as credit cards. The ongoing stresses relate to both the increased precarity of paid work for men and women – meaning women increasingly having to make significant contributions to household budgets – and the failure of hourly wages to keep up with gross domestic product per capita. In order to be able to buy the goods and services which made up our GDP, we needed ever more hours of household labour.

Older households were able to hold out for longer against these pressures, but not forever. Hence, most of the increases of labour force engagement for these households have taken place in the last thirty years.

Older Women’s Spending

What all this means is that, in the 2020s, a critical component of consumer spending is done by older households, and in particular older women. Their spending is a major source of ‘stimulus’ in the 2020s’ economy. It is already apparent that suburban cafes, for example, survive very much with the help of patronage from groups of older women.

By and large, most policymakers worldwide have now forgotten the lessons of the Great Depression of the 1930s. One of the most important lessons was that countries which had inbuilt means to keep incomeless households spending suffered much less in the peak years – the early 1930s – of that Depression. (These countries included the United Kingdom and Sweden; they contrast with France and the United States, which were still in Depression in 1939.)

France in particular could not get out of that Depression. In part because of World War One deaths and injuries, it relied very much on immigrant labour (mainly from North Africa). It also relied on female and male urban labour from people with rural connections. So, when the Depression hit, the redundant workers – having no access to benefit incomes – simply returned to either Africa or to their parents’ small farms.

Most of Aotearoa’s older women cannot emigrate if they lose their incomes. But most of them will not be able to draw on a benefit to offset their lost wages. Some are already receiving New Zealand Superannuation, and that will rise a little as the marginal tax rates on their ‘Super’ will come down. What of those under 65 who lose their incomes, noting that many employed women age 55-64 live in households which pay mortgages or rent? Most will not qualify for an MSD benefit; they will be fully reliant on their partners’ or adult children’s wages. Some, who do qualify for benefits, will face stand-downs of several weeks or months; and time engaging with MSD that would be better spent with their grandchildren or elderly parents.

One particular group of older women is those, mainly in their early sixties, who used to be able to get a ‘non-qualifying spouse Superannuation benefit’, ie if their partners were superannuitant pensioners with minimal other income. (With zero fanfare, one of the first things the Labour Government did, in October 2020, was to cancel these women’s entitlement to what was an important form of transitional income support.) These women, grandmothers in large part, are the ‘breadwinners’ in their senior households. If they lose their jobs (or their ‘roles’ as we are now supposed to say), that means a potentially catastrophic loss of household income. (We should note as an example that the New Zealand Polytechnic sector, currently undergoing significant restructuring and financial downsizing, has a particularly important portfolio of older female employees; many of these workers have substantial institutional memory, keeping their organisations functioning more than many of the younger managers appreciate.)

MSD should be focussed on helping young people to find paid work, and not having their resources logjammed by older women who would have previously had access to income support without red tape.

The Laws of Stimulus

The First Law of Holes, is ‘stop digging’. (We note that a ‘depression’ is, literally, a hole.) Finance Minister Nicola Willis is digging furiously, burying alive suffocating Kiwis.

The first law of stimulus is to stop public-sector retrenchment. That is the main single lesson from the near-forgotten Great Depression. The second law of stimulus is to have rights-based alternative sources of income that individuals of all ages can fall back on. The third law of stimulus is to stop pursuing a monetary policy that jacks-up interest rates; the ‘cost-of-living crisis’ is substantially a ‘cost of jacked-up interest rates’ crisis. (As I have already noted, debt is something that drives more people into the labour force; it’s not just the amount of debt, it’s also the cost of that debt.)

We may note that New Zealand got out of the Great Depression by adopting all three laws of stimulus. And a fourth law, by using the cheap money to embark upon a very successful ‘state housing’ program, New Zealand recovered in 1936 to 1938 with double-digit economic growth and near-zero inflation. Some of those houses, well-built, are worth a fortune now. Fletchers and other capitalists made a fortune, too; this is the kind of stimulus which would meet Simon Bridges’ business-perspective criteria. Homelessness was not acceptable to New Zealanders back then, as it seems to be now. Are we looking at a coming decade of escalating homelessness for older women?

When just about every adult is ‘in the labour force’ – unhidden or hidden – desperately needing income while employment ‘roles’ are in decline, the social stresses cannot be contained forever. Younger people may revolt, turning to the underclass-politics of the street. Older people are more likely to die unseen, as too many did in July 2022 (many denied desperately-needed second-booster vaccines) when the Covid19 pandemic really hit Aotearoa New Zealand.

Do any groups of influential people out there have the imagination and capacity to answer the call for humane economic revival? Or is it a case of those who would can’t, and those who could don’t?

————-

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The above chart shows – in red – the annual percentage increase (since 1988) in numbers employed of women aged 65-69, based on Household Labour Force Survey employment data. (And it shows, for comparison, males aged 30-34; in blue, their percentages are shown on the right-hand side of the chart. I explain below why I contrast older women with younger men.)

The employment growth of older women is particularly variable. But there are some clearly discernible patterns. To help show these, I have used “vertical gridlines” 33 months apart; 2¾ years is the persistent period of New Zealand’s trade cycle.

There are several reasons why employment may go up or down. First is simply the growth of the population for the demographic portrayed. For the latest data, the most recent women portrayed (the 2025 data point) were born around 1957 (ie born in 1957±2). Birth numbers in New Zealand peaked in the decade from 1955 to 1964; so, there will be many more women still alive in this birth cohort than in previous cohorts, and recent international migration will be low for that age group. Especially as life expectancy has been rising, population growth is a major reason for an increase in women aged 65-69 who are employed. So, growth of employment in this demographic should be well above the zero showing for the last two years.

Another reason for higher employment numbers is ‘labour force participation rates’. Charts that I posted recently from the same dataset show participation rates for women aged 65-69 having risen to 44% in 2022, the last peak of the cycle. (See my Employment in New Zealand – especially of women – at the Age Margins.) It’s unlikely that the actual participation rate has fallen since 2022; more likely it has risen in line with the trend this century (participation up from 10% to 44% of the available population); although the official participation rate has fallen. The difference between the actual and official participation rates is known as ‘hidden unemployment’)

The third reason for changing employment for ‘retirement-age women’ is the ‘added-worker effect’. This effect, highly apparent for this demographic, means that employment and actual labour force participation move essentially counter to the economic cycles (including the 33-month trade cycle). This countercyclical effect, similar to the enrolment patterns for tertiary education, is particularly apparent from 1988 to 1997. And it’s understated by the employment data, because the times when more older women want to be employed will be times of generally high unemployment. The data peaks here – eg in late 1988 and mid-1991 – reflect both the increased desire for paid work, and the reduced ability to secure paid work.

The ‘added-worker effect’ operates when other sources of household income are reduced, or when major costs such as mortgage interest or rent are high and/or rising. For this demographic there is also a ‘subtracted worker effect’, meaning that these women choose to retire whenever they can afford to retire.

It is apparent that, for the most part, the peak increases in the employment of older women follow the 33-month trade cycle. The red peaks are on or close to the charts’ vertical gridlines. There were however disruptions to the cycle caused by the 2008/09 Global Financial Crisis, and in the 2020s thanks to the Covid19 pandemic.

Males aged 30-34

The blue graph for males aged 30-34 shows the economic cycle as we would expect. This is a demographic with a very high and stable labour force participation rate. As is apparent, the blue plot is to an extent countercyclical to the red plot. The main male employment cyclical peaks were in the mid-1990s, the mid-2000s, the late 2010s, and in 2023.

We particularly note that employment growth for youngish men was weak in the late 1990s and the late 2000s. Rising young-male employment was particularly strong between 2024 and 2018, reflecting strong immigration for this demographic.

Since the present government has been in power – ie from November 2023 – employment growth for youngish men has plummeted, despite high levels of net immigration, and despite this being the baby-blip generation born in the early 1990s. Basically, the economy ‘tanked’ in 2024 and 2025.

Female birth cohorts

The second chart shows women’s employment/participation for different ‘generations’. For women born around 1957, employment peaked at just under 80% of the available population, when they were aged about 47. For the next generation, that peak is even higher, at about 83% of available women; and this is despite more women having babies later in life.

Generally, younger generations of women have had markedly higher participation rates, especially between ages 25 to 40.

Finally

Once upon a time ago we ‘worked to live’. Since the neoliberal and feminist revolutions, we have ‘lived to work’. I am not convinced that this is progress. Progress is supposed to be productivity growth; more outputs per unit of (labour and other) inputs. What we have seen is much more input – with labour inputs being shown here – yet economic growth if anything seems to have slowed.

To have a sustainable future, we should be stabilising output, while contracting inputs. Employment counts, as defined by the HLFS, are crude measures of inputs. It is perfectly possible to have more people employed, wanting and getting fewer hours per week on average. That’s not what we have here. Rather, we have more and more people desperate for employment to pay the bills, and a substantial decline in time committed to the other non-income-focused aspects of life. Women have been on the frontline of this seek-more-work play-less zeitgeist.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Quarterly Labour market data in Aotearoa New Zealand was released today. Much of the data is functionally useless, because of definitions which disguise rather than reveal important trends and turning points.

I have focussed on employment data (although the definition of ’employment’ is too generous to be optimally useful) relative to estimated populations for age groups at the younger and older margins of the ‘working age’. (For me, ‘active in the market economy’ means meeting the official definition of employment. Unlike the International Labour Organisation, I am not classing unemployed people as ‘active in the market economy’.)

The first chart focuses on older women, a group particularly impacted by recent and ongoing economic changes. Many of these people neither qualify for benefits when they become redundant; nor do they even make the official unemployment data because, compared to people aged 25 to 55, they are more often regarded as withdrawing from the labour force when they lose their jobs.

We can see that the post-1980s’ trend for all depicted age groups is one of rising ‘participation’ in the labour market; much of this is for financial reasons (eg needing to pay mortgages or rent), rather than for lifestyle or feminist reasons. Some of the change of course is linked to the increase of the age of entitlement to New Zealand Superannuation, from 60 to 65.

We see that recent (ie post-2022) data shows a flattening of the trend, or a fall against the trend for 65-69-year-old women. Most of these are actual unemployed people who are counted as ‘discouraged workers’ or ‘retired’. In reality, the financial pressures on older women to stay working are stronger than ever. 2022 represented the year of peak-grandmother-labour.

The second chart shows the trend fall in older men and women NOT employed. We note that New Zealand Superannuation – a Universal Basic Income for Seniors – incentivises people to stay working after age 65. (Australia has a substantially lower proportion of people over 65 in employment.) The trend in falling non-employment has been arrested by the greater difficulty since 2022 in finding paid work.

For younger people, the trend is for more employment and less higher-education (although many people in higher education also meet the definition of being employed). It would appear that many New Zealanders in their twenties have returned to higher education in lieu of being employed, looking to live from student allowances or student loans.

The third chart compares the teenage workforce with workers of peak working age (45-49). For peak working age we see convergence of male and female participation, despite more women in their late forties with children aged under 10 in their care.

For teenage workers, the male data fluctuates more than the female data. In this decade, more teenage males are not employed than teenage females; a clear change from the 1980s. There was a dramatic ‘flight to employment’ for teenagers after covid; a return now reversed as the job market clams shut.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

New Caledonia’s population has shrunk to 264,596 over the past six years, the latest census, conducted in April and May 2025, has revealed.

This compares to the previous census, conducted in 2019, which recorded a population of 271,400 in the French Pacific territory.

To explain the population drop of almost seven thousand (6811), Jean Philippe Grouthier, Census Chef de Mission at the French national statistical institute INSEE, said that even though the population natural balance (the difference between births and deaths during the period) was more than 11,000, the net migration balance showed a deficit of 18,000.

READ MORE

In terms of permanent departures and arrivals, earlier informal studies (based on the international Nouméa-La Tontouta airport traffic figures) already hinted at a sharp increase in residents leaving New Caledonia for good, after the destructive and deadly riots that erupted in May 2014, causing 14 dead and over 2 billion euros (NZ$3.8 billion) in damages.

The census was originally scheduled to take place in 2024, but had to be postponed due to the civil unrest.

“New Caledonia is probably less attractive than it could have been in the 2000s and 2010s years,” Grouthier told local media yesterday.

However, he stressed that the downward trend was already there at the previous 2019 census.

‘Not entirely due to riots’
During the 2014-2019 period, a net balance of around then 1000 residents had already left New Caledonia.

“It’s not as if it was something that would be entirely due to the May 2024 riots,” he said.

At the provincial level, New Caledonia’s most populated region (194,978), the Southern Province, which makes up three quarters of the population, has registered the sharpest drop (about four percent).

Meanwhile, the other two provinces (North, Loyalty Islands) have slightly gained in population over the same period, respectively +2.1 (50,947) and +1.7 percent (18,671).

The preliminary figures released yesterday are now to be processed and analysed in detail, before public release, ISEE said.

The latest population statistics are regarded as essential in order to serve as the basis for further calculation for the three provinces’ share in public aid as well as planning for upgrades or building of public infrastructure.

The latest count will also be used to organise upcoming elections, starting with municipal elections (March 2026) and provincial elections later that year.

Article by AsiaPacificReport.nz

Recently I have published charts showing how people born around 1960 are already placing huge burdens on New Zealand’s healthcare system (Death Frequencies in Aotearoa New Zealand, by Birth Year, 26 Sep 2024) and how big falls in age-specific death rates have plateaued since 2010, and may now be reversing upwards (Death Rates of Older Working Males in Aotearoa New Zealand, from the late 1970s, 17 Oct 2024).

Here I look at decennial increases in total deaths by ‘generation’, where each generation is a ten-year birth cohort centred on a zero year. Featured generations are the ‘lucky generation’ (b. circa. 1940), post-war baby-boomers (b. circa. 1950), generation Jones (b. circa. 1960), generation X (b. circa. 1970) and generation Y (b. circa. 1980).

In Table 1 above we see that in 2020, 109% more people born around 1970 died than in 2010. The main reason for the increase is that these ‘Gen-X’ people were ten years older in 2020 than in 2010. Secondary reasons could relate to the net-immigration between those years for that age cohort, or could relate to the underlying health attributes of generation-X.

(Note that the ‘+’ in the labels arises because, due to data limitations, the definition of the generations used varies slightly for each year. Thus, for 2021, Gen-X is 1966-1976.)

We see that all of the followed generations show marked increases in the increases of deaths as we progress from 2020 to 2022, with the younger age cohorts showing increased increases in 2023 as well. Generation-X is highlighted as having the biggest increases in each of these four years: 2020, 2021, 2023, 2024. This suggests underlying health issues in this generation, or greater increases in net immigration for Gen-X (compared to say Gen-J or Gen-Y), or both.

On the matter of Gen-X net immigration, we note that immigrants must undergo health checks, so it’s likely that the death rates of Gen-X immigrants since 2010 are lower than the death rates of Gen-X non-immigrants. So it’s looking like there are significantly problematic health issues being experienced by Gen-X Aotearoans.

Table 2 focusses on just February to May data. These are the months in which death numbers are generally lowest. Older people tend to die more in winter, and younger people in summer. These are mainly autumn data. We also note that Table 2 allows us to access 2024 mortality data.

This is more worrying for Gen-X, because the data show higher rates of death increase from 2022, with an especially problematic number – a 170% increase in deaths – for 2024. These data definitely suggest there’s an underlying health problem, especially for that generation. The problem may be in two parts: underlying health status (eg incidence of chronic illnesses), and increased inadequacy of healthcare (including inability to access life-saving drugs).

Table 3 focusses on July and August only, the two main months for deaths attributable to respiratory infectious diseases. As we would expect, the older generations come out ‘tops’ in 2020 and 2021. But in 2022, the year the pandemic hit in New Zealand, it’s Gen-X again which has copped the biggest increases in deaths from infectious causes. Further, in 2023, it’s the younger generations – Gen-Y as well as Gen-X – that are showing the greatest increases in winter deaths. (Lack of access to Covid19 boosters might be part of the problem here.)

So, while yesterday’s charts might have showed that life-expectancy improvements have bottomed out after 2010, these table suggest that very recent mortality data is showing definite signs that life expectancies are starting to fall, with Gen-X – born around 1970 – taking the lead in this new development.

From the point of view of funding the healthcare system, not only is the aging of the population not being properly accounted for, but also substantial swathes of the bulging generations (generations J and X) are seemingly less healthy. We remember that deaths are only the ‘tip’ of the disease ‘iceberg’; mortality increases indicate underlying morbidity increases, and it is morbidity that places the greatest demands on healthcare.

(Is there a ‘sound’ fiscal argument for expanded access to euthanasia in the coming decades?!)

*******

Data is from Statistics New Zealand, Births and deaths: Year ended June 2024.

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

This chart essentially shows the stresses that New Zealand’s public health system can expect to face. I have analysed the death data by age, covering all deaths from July 1998 to June 2024. For those years (using June years) I have looked at every age of death from 16 to 99 and every birth year from 1900 to 2022, and counted deaths by birth-year.

For death-age 95, the most frequent birth year was 1928. As we would expect, most deaths at these high ages occurred in 2022 or 2023, thanks to Covid19. Thus, birth years in the 1920s feature in the chart.

Birth years in the early 1930s don’t feature so much because of the low birth numbers in those years. With fewer people born in say 1933, then 1933 will not often feature as the most frequent birth year for any given age.

Birth years around 1950 do not feature. This is both because the classic baby boomer generation is a healthy generation, and also because there were not as many births in the decade after World War Two as there were in the following two decades. So, while classic baby boomers will place an increasing burden on the public health system, the biggest burdens will come from those born between 1955 and 1975. (Also, classic baby boomers have high levels of private health insurance; this will be less affordable for subsequent generations as they age.)

Birth years from 1955 to 1964 feature most strongly, mainly because births in New Zealand peaked in those years. This birth cohort will place massive pressure on New Zealand’s public health system. People dying since 1998 between age 37 and age 67 are most likely to have been born in the years either side of 1960.

The cohort born 1966 to 1974 will also place huge pressures on Te Whatu Ora (Health New Zealand), in part because there are likely to be very many new Aotearoans in this birth cohort. By and large, immigrants are healthier than the New Zealand born population, because their health is considered before New Zealand residency can be granted.

The late 1970s represents a ‘baby-bust’ generation, like the early 1930s. Hence these ‘Gen-Y’ people don’t feature in this chart. The frequencies for the late 1980s’ and early 1990s’ birth years reflect the ‘baby blip’ which began in 1987. Also, these birth years relate to death of young people, which, being less frequent, can also be a bit more random.

People born in 1939 turn 85 this year. From 1938, birth numbers generally increased each year until the early 1960s. The impact of an aging population on New Zealand’s public healthcare system is certainly beginning. This impact will escalate each year for at least the next 25 years. People born in 1961 will turn 85 in 2046.

By contrast, we have been lulled into complacency because the unusually small early-1930s’ birth cohort placed a substantially below-average pressure on public healthcare.

We note that death numbers are a proxy for the demand for high-intensity healthcare. People born after 1955 are already making considerable demands on Aotearoa New Zealand’s health care.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The above chart shows the votes for the principal ‘leftish’ political parties in England from 1992 to 2024. The important thing to note is that vote tallies should be rising over time in any country which has a rising population. England had had a rising population trend, yet the numbers of votes cast for the established centre-left parties have been on a falling trend.

For Labour the situation is worse than it looks. In 1992 Labour was comfortably defeated by Conservative. Yet Labour got a million more votes in 1992 than it did in 2024.

We may blame ‘apathy’ for this situation. Many more people are not voting at all. But apparent apathy is usually a symptom of something else. Ideally, when we vote we are voting for some ideal or somebody. More people vote when they perceive at least one of the options in a positive light. There is another situation which can lead to a high propensity to vote; namely if the existing government is perceived as being so bad that people will vote for whoever they must vote for in order to dismiss the government. This was the situation in England in 2024; yet even that urgency failed to galvanise voters. The total number of votes cast in England was the lowest since 2005, when Labour ‘won’ with 35% of the vote. (In 2024 Labour got 34% of the vote.)

In 2024, the total votes cast for Labour in England fell by nearly a million, after the 2019 election which was disastrous for Labour. Yet the number of seats Labour gained nearly doubled. Clearly this last distortion is a result of the ‘plurality’ voting system used in elections to the Westminster Parliament. But there’s something more important going on. The centre-left is losing favour.

The vote for the Liberal Democrats also fell in 2024, despite that party gaining a huge increase in the number of seats won. Their decline in votes is the result of what is commonly known as tactical voting; in this case it appears that about a million people who would have voted LibDem in an MMP election chose to lend their votes to Labour. (Probably more LibDem supporters than this lent their votes to Labour, because it is also clear that, where the LibDem candidate was better placed to beat the Conservative candidate, many otherwise Labour voters lent their votes to LibDem candidates.

It was this ‘efficient’ and rational vote-lending behaviour that enabled the centre-left to win so many seats. So, while, for once, ‘progressive’ voters were clever this time, the bigger story is the decline of popular support for the centre-left political agenda.

Another feature of the 2024 election is the Palestine-Gaza factor. In many traditionally Labour seats, there were ‘independent’ pro-Palestine candidates who cannibalised the Labour vote; indeed a few of these candidates won their seats.

The other important feature is the rise of the Green Party as a left-wing party winning pro-Palestine votes; especially votes of non-Muslims who are disturbed by what is currently happening in the Levant. For this see the two tables below. The Green Party may have gained ‘critical mass’, being poised to be the new left presence in British politics.

England General Election Results
Votes
	Total	Labour	Conservative	LibDem	Green	other
1992	28,148,506	9,551,910	12,796,772	5,398,293		401,531
1997	26,058,712	11,372,329	8,780,881	4,677,565	60,013	1,167,924
2001	21,870,762	9,056,824	7,705,870	4,246,853	158,173	703,042
2005	22,713,855	8,043,461	8,116,005	5,201,286	251,051	1,102,052
2010	25,085,097	7,042,398	9,931,029	6,076,189	258,954	1,776,527
2015	25,571,204	8,087,684	10,517,878	2,098,404	1,073,242	3,793,996
2017	27,165,789	11,390,099	12,379,200	2,121,810	506,969	767,711
2019	26,909,668	9,152,034	12,710,845	3,340,835	819,751	886,203
2024	24,288,122	8,365,122	6,279,411	3,199,060	1,780,226	4,664,303

England General Election Results
Seats
	Total	Labour	Conservative	LibDem	Green	other
1992	524	195	319	10
1997	529	329	165	34		1
2001	529	323	165	40		1
2005	529	286	194	47		2
2010	533	191	298	43	1
2015	533	206	319	6	1	1
2017	533	227	297	8	1
2019	533	180	345	7	1
2024	543	348	116	65	4	10

Watching that election on UK Sky TV (live on You Tube), one commentator repeatedly mentioned the “efficiency” of Labour, meaning that Labour won many seats on small margins. This so-called efficiency will make Labour very vulnerable in the next election, which, luckily for them, may not be until 2029.

Unless Labour performs exceptionally well, the votes lent to Labour will return to their LibDem homes.

What about the votes Labour lost to Independents and Greens in safe Labour seats? And the votes, Labour lent to winning (and near-winning) LibDem candidates. They are most likely to stay with the Liberal Democrats who will need these votes to fend off Conservative candidates.

The Tories

What of the ‘Tory’ Conservatives? They clearly got trounced; their vote count fell by more than 50% in the 2024 election. They may or may not get votes from people who voted Reform, the biggest of the ‘other’ parties in 2024. A useful strategy for them could be to cultivate the large conservative Muslim vote. A significantly higher proportion of voters in England are now Muslims; that proportion will only grow as Muslim households continue to have more children than the national average. And, Islam is a very conservative religion.

There is a natural fit here, going forward. Former Chancellor of the Exchequer Nadhim Zahawi was born in Iraq and is “thought to be a Muslim”. Likewise, another former Conservative Chancellor, Sajid Javid – born to Pakistani parents – “still identifies as being a Muslim”. If the Tories wish to be relevant in England’s future, they will need to adopt a wider political vision that is attractive to non-radical Muslims as well as to conservative people of other faiths. Otherwise, the future of the Right in England may fall to the new Reform Party; such a change has already happened in France.

So, I am predicting that – in 2029, or before – the LibDems may come through the middle, just as Emmanuel Macron’s party did in France in 2017, leaving both Labour and Conservative to play the role of small ‘legacy parties’. Labour’s ‘landslide’ is likely to accelerate, but in the wrong way; indeed, a landslide is actually a disaster.

Note

In this chart and text, I have looked at England only, which is the core of the United Kingdom, but not its entirety. This is because, especially in Scotland and Northern Ireland, other parties play significant roles. In Scotland in 2024, the big story was the crash of the Scottish National Party (SNP). Labour was a beneficiary of that crash. But it is likely that votes lent to Labour by regular SNP voters will not stay with Labour.

______________

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

In this article, I will simply present the numbers for the key performance indicators (taken from tradingeconomics.com). And then I’ll give an update on GDP (gross domestic product) data, given today’s data release from Statistics New Zealand, and with the definition of ‘recession’ in mind.

An important part of the context is that for many years, the principal economic commentary about Japan we hear in Aotearoa New Zealand is that the Japanese economy is “sluggish” and has been so for many years.

Interest Rate (OCR equivalent)
Dec-2014	3.50%	0.00%
Dec-2018	1.75%	-0.10%
Dec-2022	4.25%	-0.10%
Dec-2023	5.50%	-0.10%

Inflation Rate (CPI)
Dec-2014	0.8%	2.4%
Dec-2018	1.9%	0.3%
Dec-2022	7.2%	4.0%
Dec-2023	4.7%	2.6%

Unemployment Rate
Dec-2014	5.5%	3.4%
Dec-2018	4.3%	2.4%
Dec-2022	3.4%	2.5%
Dec-2023	4.0%	2.5%

Current Account as % of GDP
Dec-2014	-3.4%	0.8%
Dec-2018	-4.2%	3.5%
Dec-2022	-8.9%	1.8%
Dec-2023	-6.9%	*2.9%
		* estimate

Government Budget as % of GDP
Dec-2014	-1.2%	-5.9%
Dec-2018	1.9%	-3.4%
Dec-2022	-2.7%	-6.4%
Dec-2023	*-1.6%	*-5.0%
		* estimate

Government Debt as % of GDP
Dec-2014	25.3%	233.5%
Dec-2018	19.4%	232.3%
Dec-2022	35.9%	263.9%
Dec-2023	*38.4%	*265.9%
		* estimate

Locating Recessions

The table below highlights recessions using two different definitions, for aggregate GDP growth and for GDP growth per person in the population.

The two definitions are essentially those given on 20 March, but with small modifications. The definition using Q-to-Q data (quarter to quarter, seasonally adjusted), is when two subsequent quarters have a lower GDP than a given quarter. For example, in New Zealand in 2021, both the December and September GDP values were lower than the June value.

The second definition applies to Y-to-Y data, meaning when a quarter is compared to the same quarter of the previous year. Any negative value counts as a recession. Additionally, that recession may be backdated; an example is New Zealand in June 2023, where the GDP for that quarter is less than the GDP for September 2022.

It may be noted that there are more recession quarters in New Zealand when per capita data are used, because of a general pattern of rising population in New Zealand. This is particularly relevant in 2023. Japan on the other hand has fewer quarters in recession when the data is adjusted for population change; this is because Japan’s population is falling, and that rate of decline has sped up in recent years.

Takeaways

While all of the data presented is important for the reflection of policymakers, I would particularly like to emphasise the importance of interest rate settings (monetary policy), and government budget settings (fiscal policy). In this regard, and given the ongoing political attention given in New Zealand to government debt, that New Zealand is about number 140 in the global government debt league, whereas Japan is number 2.

And we should note that the current account data indicate that New Zealand as a country – ie considering the country as a whole, not just the government – is a significant debtor country whereas Japan is not. Desirable current account numbers are within three percent (plus or minus) of GDP. (Interestingly, New Zealand’s current account data for the December 2023 quarter were released on 20 March, but to much less fanfare than the GDP data release today. There would appear to be a degree of media self-censorship in this regard.)

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Generally, more people die in winter. Not surprising, though some years have significantly more deaths than others, and the timing of ‘peak death’ each year varies between the wintery months. These charts show the deaths, determined from weekly data, of people named Smith, New Zealand’s most common surname last century.

The numbers shown are nine-week moving totals, meaning that for the last week of July the data runs from the beginning of July to the end of August. The next datapoint drops the first week of July, and includes the first week of September. (This method addresses the randomness of death, and the randomness associated with the Smith sample.)

Secular Trend?

It is somewhat surprising that the numbers of deaths in 1973 were not much higher than in 1950. The population of New Zealand in 1973 was 3.0 million; in 1950 it was 1.9 million. More people should mean more deaths. But the age structures were quite different. In 1950, there were relatively many older people – thanks to the 1870 to 1895 baby boom. In 1973, there were fewer older men thanks to both World War One, and to the deceleration in birth numbers from the 1890s. With the partial exception of the early 1920s, that reduced birth rate lasted from around 1900 to 1945; though there was variation, with say the early 1940s having many more births than the early 1930s.

But, what goes around comes around. There was another baby boom from 1945 to 1975; a boom that is only just starting, in the 2020s, to markedly influence death tallies. So, as annual death numbers have been only on a slow incline in the lifetimes of those alive today, annual death numbers are set to increase dramatically. Just as individuals die, so do generations. And big generations die bigly.

(We also note that, in the 1940s and 1950s, infant mortality was much higher in New Zealand than in the 1970s.)

Mortality Peaks

The higher peaks in these charts can be attributed to influenza outbreaks. In addition, the winter seasonal highs are linked to the set of viruses – including coronaviruses – which we collectively know as the ‘common cold’.

Superficially, these charts suggest that ‘the flu’ and ‘the common cold’ are New Zealand’s grimmest reapers; are, together, New Zealand’s biggest public health nuisance. Further, the peaks in these charts seem to be getting higher relative to the troughs in the more recent data. Should this be a matter of concern? Didn’t we, by the 1970s, reach a state of hubris about infectious diseases?

Old Age

Death and taxes are (allegedly) the two principal certainties of life. If we don’t die of one thing, we die of something else. So, an increase of deaths triggered by ‘minor’ respiratory viruses can be explained, mainly, by a relative decrease in deaths from other causes such as heart disease and cancer.

And there may be more to it than that. The seasonal circulation of non-novel respiratory viruses may represent a kind of natural vaccination programme. So, at least for otherwise healthy working-age (and younger) people, the presence of these viruses in our temperate ecosystems may be contributing to our increased longevity. Less smoking and sugar, combined with more (not less) exposure to respiratory viruses, may be the essence of why life expectancies have risen in recent decades.

If so, then, the presence of highly seasonalised death patterns may represent a collective solution rather than a collective problem. On balance, influenza may be our friend, not our enemy. It may determine the timing of death in old-age rather than be a significant cause of premature death.

(Tactfully, Queen Elizabeth’s death certificate in 2022 simply attributed her death to ‘old age’. Old Age is a real thing, and not an expression of casual ageism. Indeed, ‘old age’ was the most important and truthful part of her death story; though, as is usual, a single attribution is not the whole story of a person’s death.)

Dry Tinder

Outbreaks of influenza (and other respiratory) viruses work like forest fires. Thus, after years with relatively few winter seasonal deaths, there is a build-up of ‘fuel’ meaning that there will soon be a year or two of higher numbers of seasonal deaths. Followed by years of below-average winter deaths. This is a normal pattern. When there is a large build-up of people of advanced age, there will be more deaths from old age. That’s the normal cycle of life. How do people die of old age? More often than not, such deaths are triggered by a seasonal infection.The aim of public health policy is to maximise the numbers of people who die of old age; minimising the numbers who die prematurely.

These charts, to a large extent, represent deaths due to old age. They also indicate years of more virulent strains of influenza.

The Charts

In the purple 1946-1955 chart, we see 1950 and 1953 as the years of elevated winter deaths, suggesting more dangerous influenza strains. We also notice secondary death peaks in late spring, early summer. In 1949 and 1952 these secondary peaks were higher than the primary winter peak for that year. Presumably, the end of the year is a time when people circulate more; and there will be more vulnerable people if the winter death tally was unusually low.

Looking at the red 1955-1964 chart we see 1956 looking much like 1950, suggesting two low-mortality years would be followed by a higher mortality year, presumably the ‘dry tinder’ effect.

1957 and 1958 were the years of the ‘Asian Flu’ pandemic (a novel strain of influenza), and it shows in the New Zealand data for both years; higher death tallies in years which would otherwise have had significantly fewer deaths. Many of these additional deaths will be of people who would otherwise have lived a few years longer. (Unlike the extremely lethal 1918 strain, most non-elderly people with good general health seem to have weathered this pandemic OK.)

As is normal after a respiratory pandemic (and this is certainly true in Eastern Europe after Covid19, where public health measures substantially subsided in the latter part of 2020), the death tallies for the next couple of years (1959, 1960) is significantly down. 1961 and 1964 were higher winter mortality years, as per the three-year pattern. (1963 had a sharp mortality peak, probably a nasty flu strain, followed by unusually low spring mortality.)

Looking at the blue 1964-1973 chart, 1968 to 1970 reflects the ‘Hong Kong Flu’ pandemic. Whereas the 1957 influenza strain was first reported around January of 1957, the 1968 pandemic strain was first reported in the middle of that year. There was no sign of it in New Zealand in 1968, or in early 1969. Then, in mid-1969, with a mix of ‘dry tinder’ and a lethal influenza strain, there was a longer than usual mortality peak. Then, after a short pause, the pandemic really hit in December, and lasted until August 1970. Like the 1918 pandemic influenza peak, and some of the Covid19 peaks, this was a summer shock.

1971 and 1972 were also high mortality years, suggesting that many who died from influenza in the early months of 1970 had been of working age rather than old age. There were still many frail old people in the population after 1970; people born during the first baby-boom era.

By 1973, we start to get the impact of diminishing relative numbers of older people; a combination of Word War and falling birth rates around the year 1900.

Prognostication

These charts show pandemic and substantial epidemic influenza outbreaks in New Zealand. And they show how ‘old age’ deaths follow a seasonal pattern; commonly triggered by a respiratory virus which would be weathered by the vast majority of people who did not have the characteristics of old age. These viruses are part of the wider ecosystem of which humans are very much a part. Further, the ecosystem of seasonal viruses is maintained by periodic appearances of virulent novel viruses.

There is no reason to believe that life expectancies could be raised by taking public health measures to eliminate influenza and ‘cold’ viruses. Rather, these viruses fine-tune our immune systems, and without that fine-tuning, average life-expectancy would probably fall. Indeed, one cannot imagine the possibility of healthy populations in the crowded metropolises of the world without regular exposures to non-lethal respiratory viruses; exposures tantamount to natural vaccination.

Note

From 1974, not all historical deaths can be accessed online. The rule is that, for today (13 February 2024), only deaths of people born on or before 13 February 1944 will be accessible. This means that 1974 data will not be comparable with 1973 data, because it will miss about a half of infant deaths. We should also note, however, that by 1974 infant mortality rates were substantially lower than they were in the 1940s and 1950s; meaning that late 1970s’ Smith data will remain broadly comparable. A substantial majority of the ‘Smithometer’ Smiths after 1973 will be older people.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.