Reading the Room: What 'Work From Home' Searches Reveal About Economic Shock and Behaviour in South Africa

Mon, 01 Jun 2026 00:00:00 +0000

This post is a data analysis essay. The 2026 Google Trends estimates are reconstructed from documented patterns and cross-referenced with independent analyses. Sources are listed at the end.

Two spikes on a single index. COVID peaked at 100. The Iran war reached ~62. Same signal, different mechanism.

If you work in a South African office, you have now lived through two distinct “work from home” moments and there is a reasonable chance you did not fully notice that they were structurally different experiences or perhaps you did?.

The first was March 2020. President Ramaphosa appeared on television. By the following morning, you were either setting up a laptop at your kitchen table or discovering, abruptly, that your job was one of the many that could not survive that particular translation. You remember the sudden strangeness of video calls, colleagues’ bookshelves, children appearing in doorways, and the unsettling intimacy of a world rearranging itself overnight.

The second is happening now, more quietly. There has been no lockdown announcement, no state of disaster. But since March 2026, when Brent crude crossed $100 per barrel in the wake of the Iran war and South African petrol prices rose by more than R6 per litre, the arithmetic of commuting shifted enough that many employers — without fanfare, without a policy document, quietly reopened the work-from-home option. Not as a public health measure. As a cost measure.

Same behavioural signal. Different mechanism. And as an economist studying how people respond to income shocks, I find this comparison difficult to look away from.

Google Trends as a Behavioural Observatory

Before getting into the comparison, it is worth being precise about what we are actually measuring.

When someone types “work from home” into Google, they are not necessarily working from home. They are signalling consideration, they are calculating, or at least entertaining the question. In the language of revealed preference theory, a search query is a low-cost signal of deliberation: closer to intent than action, but measurable and near-real-time in a way that survey data is not.

Google Trends normalizes search interest on a scale of 0 to 100, where 100 represents peak popularity of the term in the selected period and geography. It is not an absolute count of searches; it is a proportional signal. Its value for economic analysis is not precision, it lacks that, but timeliness and behavioural directness. Survey data on workplace adaptation takes months to collect and publish. Search data reflects what millions of people are thinking about the morning after a policy announcement, a price spike, or a pandemic declaration.

For academic purposes, high-frequency search data has been used as a proxy for unemployment expectations, housing sentiment, consumer confidence, and pandemic behavioural response. It is noisy, but directionally reliable, and when two different crises produce two different spikes on the same index, the comparison is worth taking seriously.

What does the “work from home” index look like from January 2020 to May 2026? The data shows two clear spikes above a stable long-run baseline:

The first spike: 100 in the week of 16–23 March 2020. The all-time peak.

The second spike: approximately 62 in March 2026. The largest reading in four years.

The 2020 Shock: Preference, Fear, and the Overnight Migration

The behavioural economics literature distinguishes between price shocks, which change the relative cost of options without changing what people fundamentally want, and preference shocks, which change the underlying utility ordering itself.

COVID-19 was primarily a preference shock. The pandemic did not make the office more expensive in rand terms; it made the office feel existentially threatening. The commute carried the possibility of infection; the open-plan floor carried it further. For millions of workers, the ranking of work environments changed overnight from “office is the default, home is inconvenient” to “office is dangerous, home is the rational choice.”

But COVID-19 was simultaneously a mandate, a legal intervention that removed the choice entirely for most of the formal sector. South Africa’s Level 5 lockdown (26 March - 30 April 2020) did not ask people to consider working from home. It ordered it. This is a methodologically important point: the March 2020 WFH spike is not cleanly interpretable as voluntary behavioural response. It is the superposition of changed preferences and legal compulsion, and separating the two empirically is an interesting task.

What the google trends data does tell us: WFH search interest peaked at 100 the week the pandemic was declared globally — before the Level 5 announcement on 23 March. This means the preference shift preceded the mandate. People were already searching for alternatives before Ramaphosa made them legally compulsory. The decree reinforced and operationalized a behavioural shift already in motion.

By Level 4 (May 2020), the index had declined to roughly 75. By June, 50. A new long-run equilibrium, partially remote, partially hybrid — was being negotiated in real time between employers and workers. The speed of the structural change was extraordinary: in 2019, just 1.2% of South African job postings advertised remote work as an option, one of the lowest rates in the world. By mid-2020, close to 50% of formal-sector employers had some form of remote arrangement. A structural shift that might have taken a decade was compressed into six weeks.

By 2023, with return-to-office pressure mounting, the share of employers offering remote options had retreated to around 19%. But the baseline had permanently reset. The index never returned to its pre-2020 level of roughly 6; it stabilized at 15–16. The COVID shock left a structural residue.

The 2026 Shock: Price, Calculation, and the Quiet Renegotiation

The 2026 shock operates through an entirely different mechanism.

On 28 February 2026, the United States and Israel launched coordinated strikes on Iran. Within days, the Strait of Hormuz, through which approximately 20% of the world’s seaborne oil trade passes daily, was effectively closed to normal tanker traffic. South Africa, which imports 100% of its crude oil and over 80% of its refined petroleum products, was directly in the exposure pathway. By mid-March, Brent crude had surged from roughly $69 to above $100 per barrel. South African petrol prices rose by more than R4 per litre; diesel by more than R6 in some areas (even after the government cut the fuel levy by R3).

For a commuter travelling 40km each way in a mid-range vehicle, this translated to roughly R600–R900 in additional monthly fuel costs compared to February’s baseline, a meaningful real income shock against a backdrop of 32.7% unemployment and stretched household budgets. For a minibus Taxi or e-hailing service recording an average of 5500 KM a month, this translated to an extra R2000-R4000 in additional monthly fuel costs.

The International Energy Agency responded with a 10-point plan to reduce global oil demand. Item one: encourage working from home. Governments across Asia — South Korea, the Philippines, Malaysia — mandated WFH for civil servants. South African employers and employees alike, without a mandate, began their own quiet calculations.

The WFH Google Trends index rose to approximately 62 in March 2026.

This is analytically interesting precisely because the mechanism is different in every important respect from 2020. There is no health risk. No lockdown (or perhaps only a slight chance of it). No legal obligation. The office is physically accessible. The driver of the behavioural shift has been purely economic: the cost of commuting has risen enough to shift the break-even calculation on remote work for those for whom remote work is feasible.

In behavioural economics terms, this accounts for price effect. When the relative cost of an activity increases, demand for substitutes rises, provided those substitutes are available and the switch is feasible. Work from home is the substitute good for commuting. The Iran war raised the price of commuting; the search index rose in response.

The fact that the index reached 62, not 100, tells us something important about the magnitude of each mechanism. The COVID shock was both a preference shift and a legal mandate — two forces pulling in the same direction simultaneously. The 2026 shock is a price effect operating alone, and only on those for whom the WFH option is technically and contractually available. It produced roughly 62% of the behavioural intensity of COVID. That is not a trivial response; it is a meaningful signal that the price elasticity of WFH behaviour, at this cost level, is real.

The Shape of the Response: Decay Curves Matter as Much as Peaks

The magnitude comparison (100 versus 62) is the easy headline. The more analytically instructive question is about the shape of the response over time.

When behavioural change is driven by a preference shock (COVID), the decay rate tends to be slow. Risk perception is sticky. People do not quickly un-learn that they were afraid of an office. The COVID index remained above 30 for nearly two years - through Delta, Omicron, rotating lockdown levels, and eventually the state of disaster being lifted in April 2022. Habit formation, employer policy lock-in, and a re-evaluated sense of what work could look like all contributed to a structural ratchet effect.

When behavioural change is driven by a price shock (Iran war), economic theory predicts steeper mean-reversion once the price stimulus moderates. The ceasefire was announced on 7 April 2026. Hormuz tanker traffic began slowly recovering. Oil prices partially retreated. The WFH index has declined from its March peak to roughly 48 in April, 42 in May. The decay curve is visibly steeper than COVID’s first three months.

This distinction has practical implications that extend beyond the index itself. For companies currently making decisions about office space leases, remote work policies, and flexible work entitlements, the question is not just “how high did the spike go?” but “is this a structural shift or a cyclical response?” COVID was structural. The 2026 shock, based on the decay curve so far, looks more cyclical, unless it catalyses a second-order change in employer norms, as COVID did, in which case the structural residue will only be visible 12–18 months from now.

Beyond Consideration: The Equipment Signal

The “work from home” index measures deliberation — whether people are thinking about working remotely. But thinking about something and committing to it are different behavioural states, and the gap between them turns out to be analytically useful.

Consider what happens after someone decides to work from home seriously. They search for “external monitor.” They look up “ergonomic chair.” They compare webcams. These are not impulsive purchases; they are investments with a depreciation horizon. You only buy a proper home-office monitor if you expect to need it for long enough to justify the cost. The searches for these items are therefore a commitment signal — evidence not just of WFH consideration, but of permanence expectation.

Building a composite of Google Trends interest for “external monitor,” “webcam,” and “ergonomic chair” alongside the main WFH consideration index reveals something striking about the two shocks:

During COVID, the equipment composite spiked to approximately 72 but it peaked one month after the WFH consideration index. People searched “work from home” in March 2020, then ordered monitors and chairs in April. The lag is itself informative: it reflects the deliberate setup of a home working environment that was expected to last. And it did. The commitment ratio (incremental equipment signal per unit of consideration above baseline) was approximately 0.53 during COVID. For every point of WFH deliberation above the pre-crisis norm, you got half a point of equipment investment.

During the Iran war, the consideration index rose to 62. The equipment composite barely moved to roughly 28, against a baseline of 22. The commitment ratio: approximately 0.13. About four times lower than COVID.

Two explanations for this gap, and both are probably true simultaneously:

The capital stock effect. Many South African formal-sector workers already own the equipment bought in 2020 and still in use or sitting in a cupboard. The marginal cost of re-activating a WFH setup in 2026 is close to zero for those who invested during COVID. No new monitor needed. The absence of an equipment spike is not evidence of low commitment; it is evidence that COVID pre-funded the infrastructure for future shocks.

The permanence expectation effect. Workers and the employers implicitly signalling WFH permission appear to regard the Iran war WFH moment as more temporary. A ceasefire was already announced within six weeks of the shock’s peak. Fuel prices are retreating. The behavioural logic: why buy a R4,000 monitor for a situation that may resolve by the time it ships? The low commitment ratio is consistent with rational expectations of short duration.

Taken together, the commitment index reframes what the 2026 WFH spike actually means. It may not be a second wave of structural change, it is likely people cashing in on an option that COVID’s investment created. The 2020 shock built the infrastructure; the 2026 shock is merely making use of it.

Why I Am Watching This

I should be transparent about where this interest comes from professionally.

My PhD, completed at Stellenbosch University earlier this year under the supervision of , examined how Apartheid’s distortions of the South African labour market — the legal restrictions on where people could work, what jobs they could hold, where they could live — imposed measurable costs on aggregate economic productivity. The central mechanism was misallocation: workers legally prevented from reaching their highest-productivity employment matches. One chapter used upgrades to South Africa’s railway network as a natural experiment: when rail lines were upgraded and commute times shortened, workers could reach better-matched firms. Aggregate manufacturing productivity improved. The commute, it turned out, was doing economic work beyond simply moving people.

That framing — commuting infrastructure as a determinant of labour market allocation efficiency — is directly relevant to the WFH question. When working from home becomes either mandatory or economically rational, the commuting constraint is removed for a subset of workers. The resulting allocation changes. Which workers can substitute WFH for commuting, and which cannot, is fundamentally a question about which workers are constrained by geography, infrastructure, and job type — and which are not. South Africa’s structural answer to that question is stark: roughly 60% of the employed workforce cannot work from home regardless of what happens to health or fuel prices. The WFH index, however high it climbs, captures only the remaining formal sector.

In my current research role at — South Africa’s largest dedicated financier of the minibus taxi industry, with close to 50% of taxis on its finance and insurance books — I spend my days studying a related but inverse problem. The taxi industry carries 65–70% of South Africa’s public transport trips. It is the infrastructure that makes commuting possible for the majority of workers who have never had the WFH option. When the search index rises — when formal-sector workers shift to remote arrangements, it signals a reduction in demand on that infrastructure. Taxi operators carrying fewer passengers to offices they are not attending face the same income shock that my research is designed to study.

The Iran war has, in a sense, handed us a large, exogenous income shock to observe in real time alongside the randomised controlled trials we are running on loan restructuring and repayment behaviour at Mobalyz. I would not have chosen to run this particular experiment. But the data is, as these things go, extraordinary.

What the Index Cannot Tell You

Intellectual honesty requires flagging the limitations clearly.

First, Google Trends is a proxy for search intent, not behavioural outcome. The gap between “searching for WFH tips” and “actually working from home with employer approval” is not captured here and is likely substantial — particularly for the 2026 shock, where employer policy is the binding constraint for many individuals. High search interest does not mean high WFH adoption; it means high deliberation about WFH adoption.

Second, the South African search population is not a representative sample of the South African labour force. Internet penetration, while improving, remains skewed toward urban, higher-income, formal-sector workers. The index systematically over-represents the workers most likely to have WFH options, and under-represents the majority who do not.

Third, the 2026 index values in this post are reconstructed estimates based on documented search patterns and cross-referenced with independent analyses. They are directionally robust but should be treated as indicative rather than precise. You can verify the shape and magnitude directly on .

Fourth, the two shocks occur in different economic contexts — South Africa in 2026 is not South Africa in 2020. Unemployment has remained stubbornly above 32%, but a hybrid work norm already exists (a legacy of COVID), the electricity crisis has partially improved, and the institutional memory of WFH as a feasible arrangement is present in ways it was not in 2019. The behavioral response to a price shock in 2026 operates within a different set of constraints and employer norms than the response to a health shock in 2020. This complicates any clean causal comparison.

A Question for You

If you are reading this from a South African office — or from your home office, which is itself a data point — you have lived through both of these moments. You know what it felt like to be asked, or ordered, to shift to WFH in 2020. And you may have noticed, in recent months, whether the conversation in your organisation has reopened.

I am genuinely curious about how this has played out in practice, and I would welcome responses:

Did your employer reopen WFH or hybrid arrangements in response to fuel costs this year — and if so, was it formal policy, informal tolerance, or individual manager discretion?
For those who commute by private car: has the fuel cost calculation visibly changed your own decisions about when it is worth driving in?
Has your organisation factored WFH into any formal cost-reduction conversation in 2026?
For researchers reading this: is there a better high-frequency behavioural proxy for formal-sector commuting decisions that I should be using alongside search data?

The comparison between these two shocks is still playing out. The 2026 decay curve is only three months old. Whether it levels off at a new, higher baseline — as COVID did — or returns cleanly to 15 — as a pure price effect would predict — will tell us something important about whether the Iran war has shifted the long-run equilibrium of work arrangements in South Africa, or merely produced a temporary blip.

I’ll be watching the data. If any of this resonates, or if your experience contradicts it, I’d genuinely welcome the conversation — in the comments, on , or by email.

Sources

#	Source	Note
1	Google Trends	“work from home,” global + South Africa, Jan 2020–May 2026
2	IEA, 10 Point Plan to Cut Oil Demand	April 2026; WFH as explicit demand-reduction measure
3	Global Workplace Analytics, Telecommuting Statistics	Pre- and post-pandemic WFH adoption rates
4	Michael Page Africa / IntechOpen	SA remote work adoption survey data, 2020–2023
5	PNAS, The global persistence of work from home	Cross-country WFH persistence data through 2026
6	Wikipedia, Economic impact of the 2026 Iran war	Brent crude price timeline; Hormuz closure data
7	Statistics South Africa, National Household Travel Survey	Commuter mode share; household transport spend
8	Business Day / Daily Maverick	SA fuel price and economic impact reporting, March–May 2026
9	Ngalande, T. (2026). Growth, Productivity and Labour Misallocation: Apartheid South Africa. PhD Thesis, Stellenbosch University	Railways and commuting chapter
10	Statista, COVID-19 impact on remote working Google searches	2020 benchmarking and peak-period documentation
11	Google Trends — composite commitment index	“external monitor,” “webcam,” “ergonomic chair” — simple average; Jan 2020–May 2026

Timothy Ngalande is Researcher at and a research affiliate of LEAP (Laboratory of Economics of Africa’s Past) at Stellenbosch University. His research focuses on behavioral finance, labour markets, and transport economics in Southern Africa.

COVID-19 | Timothy Ngalande