Household Compensation Costs: Reading National Budgets and Poverty Lines

1. Match the country and objective before interpreting compensation costs

The United Nations Development Programme (UNDP) released the report on 11 October 2026.[1] Turning the figures into policy judgments requires asking whose budget the aggregate cost represents, and what must be restored before support is considered complete. A single international aggregate ratio cannot directly establish a particular country’s burden or households’ recovery. Comparisons become meaningful only after fixing the size of each economy and the state that protection is intended to preserve.

For a government, for example, the cost relative to the size of its own economy is not the only relevant consideration. It must also know the currency in which that cost is measured and which losses of which households it is meant to cover. A similarly named measure in another country cannot be ranked by expenditure alone if its recipients or the reference date for income differ. The claim that a small global burden is also small in an individual country, and the claim that lower expenditure achieves the same outcome, each require different evidence.

The same issue arises for households. Moving slightly above a threshold and regaining the ability to buy the food and fuel affordable before a price increase are different destinations. The number of people below a line captures one aspect of society. It cannot by itself measure additional losses among people already below the line, or reduced purchases among those who remain above it. Evaluating support through headcounts alone can obscure both protection that leaves the count unchanged and deterioration that also leaves it unchanged.

These distinctions matter for international markets. Underestimating a government’s funding requirement changes assessments of fiscal capacity; overestimating household recovery changes forecasts for food, retail, transport and other demand. Conversely, overstating a crisis by generalising aggregate conditions too far can also distort financing and business plans. The starting point is not to reject statistics, but to identify the questions they answer and test, step by step, how they connect to national budgets and household living standards. The route from prices to household and business payments is also explained in Separate international energy prices from bills (Free guide).

2. SG Group’s view: measure the average burden and household recovery separately

SG Group uses the aggregate ratio as a starting point while examining the distribution of national burdens and the endpoint of compensation separately. The first question is what survives, and what disappears, when costs across economies of different sizes are combined. The second is the state to which spending is meant to return a household. Without fixing both, a comparison cannot distinguish a lower cost caused by a change in coverage, a change in the level of protection, or an improvement in programme operation.

The first analytical axis is the mechanism of aggregation. When national costs and gross domestic product (GDP) are expressed in compatible units, the ratio of their totals is a GDP-weighted average of national ratios. Working backwards from that average to national ratios requires information beyond the average. Substituting a simple average or population shares changes the relative positions of larger and smaller economies and produces a different statistic. This is not merely a technical detail: it underpins discussions about allocating an international estimate of support needs across countries.

The second axis is how household conditions are measured. The number below a line at one date, the number crossing it since an earlier date, and the purchasing power lost relative to an earlier position contain different information. This distinction does not rank the value of individuals. It allows the same household records to show how far a headcount-reduction objective captures deterioration in living conditions, and how an objective of replacing lost income appears in headcounts. The language of an objective must correspond to the quantity observed.

One objection is that increasingly precise conditions slow policy action. But establishing a basis for comparison is different from doing nothing until every fact is available. Even a provisional decision can narrow the next evidence request by distinguishing unknown national amounts from an unsettled definition of the target group. If evidence shows that actual national burdens are close to the aggregate ratio and that household endpoints coincide across policies, many elaborate distinctions can be set aside. That possibility belongs in the test too.

3. Check what the underlying documents count when comparing costs

When two cost documents disagree, first check whether they count the same thing. If a government document uses households and another uses people, comparing totals requires converting to a common population unit even when unit costs look similar. Payment counts and recipient counts also differ: instalments can cause one person to appear several times in payment records. Such inconsistencies in units are not evidence that policy effects differ. Matching the objects counted, rather than relying on document titles, is the first condition for comparing costs.

One way to take the comparison further is to investigate what changes while the objective stays fixed. With records for the same household, varying one at a time whether prices, quantities purchased or income are held constant helps identify sensitive inputs. A calculation that changes prices alone differs in meaning from one that also changes the purchasing basket; that is not the same distinction as changing coverage. Selecting a single smaller result without recording the comparison conditions makes it impossible to test whether the choice favoured a preferred conclusion.

Evaluating policy effects with observed data also requires specifying how the comparison group was selected. If supported and unsupported households differed beforehand, their subsequent gap cannot all be attributed to support. Analysts need ways to separate its causes, such as finding groups with similar prior trends and purchasing baskets. If unmatched characteristics remain, the assumptions on which the conclusion depends must be stated. It is important to distinguish a simple before-and-after comparison from one that estimates what would have happened without support.

A further advantage of a research design is that the comparison period and main indicators can be fixed before the results are known. Selecting only the products or regions with large improvements afterwards can overstate the overall effect. Reporting only an average, however, may hide severe deterioration in a minority. Showing the overall result alongside predetermined breakdowns, and explaining why observations were excluded, makes it easier for readers to assess whether data selection has manufactured the conclusion.

4. What weights national ratios in an aggregate GDP ratio?

To examine aggregation, consider hypothetical national data already converted to the same period, price basis and common unit. Let country i’s compensation cost be Cᵢ and its GDP be Gᵢ, with GDP positive. Its cost ratio is rᵢ=Cᵢ/Gᵢ. The aggregate ratio R, obtained by dividing total costs by total GDP, equals the average of the national rᵢ weighted by their shares of GDP. This is an identity under these definitions, not a replication of a particular report’s calculation.

The weight is neither the country’s share of population nor the reciprocal of the number of countries. Because the denominator aggregates GDP, each country’s contribution to that denominator determines its weight. For the same change in a national cost ratio, a country with larger GDP has a greater effect on the aggregate; a smaller economy has less influence. The aggregate can remain stable while burdens change sharply in some small economies. GDP size does not itself determine the severity of a burden or priority for support.

This property helps test national explanations based on international averages. To establish whether a country lies above the average, its own rᵢ must be compared with it. Allocating total costs by population and dividing the resulting amount by national GDP is a new allocation assumption, not a reconstruction of the original country estimates. A government can choose population-based allocation as policy, but that choice requires its own objective and rationale. It does not follow automatically from the statistical calculation.

Decomposing a weighted average helps explain why it changed. A change in national ratios, in GDP shares, or in the set of countries covered has a different interpretation. The aggregate and some national indicators can move in opposite directions without either necessarily being wrong. Aligning periods and coverage, and examining weights separately from values, can explain such movements.

5. An aggregate value does not uniquely recover national burdens

Aggregation compresses information into a single number, which is why national ratios cannot generally be reconstructed from it. Consider two countries with a positive total cost and positive GDP weights. Even if both weights are known, only the weighted total is given. Raising one ratio slightly and lowering the other by an amount adjusted for its weight can leave the total unchanged. Different combinations remain possible within bounds that avoid inappropriate values such as negative costs. Finding one combination consistent with the aggregate therefore does not prove that it represents reality.

In symbols, with weights wₐ and wᵦ for countries A and B, increasing A’s ratio by δ and reducing B’s by wₐδ/wᵦ causes the changes in the aggregate to cancel. Here δ is not an estimate of an actual change in burden: it identifies another combination yielding the same total. The example does not claim that a particular country faces a high burden. It shows that the aggregate and weights alone cannot identify where the burden is concentrated. If all costs are nonnegative and the aggregate ratio is zero, however, every national cost ratio must also be zero; that case is outside this example of non-uniqueness.

There is a priority order for evidence that closes this gap. National cost amounts or ratios allow direct comparison; knowing values for at least some countries narrows the possibilities for the remainder. Assuming identical ratios across all countries is another option, but it should be identified as a convenient computational assumption. Applying the overall average to a country does not test whether that assumption is reasonable. This is also why differences in import structures or household purchasing baskets merit investigation.

When national dispersion is unknown, placing an arbitrary range around the average is not an appropriate representation of uncertainty. A range needs support from national observations, model sensitivity analysis or explicit assumptions. Showing that several allocations can produce the same aggregate does not establish their probabilities. For investors and businesses, the practical question is what additional information would identify a conclusion before using the average to assert a particular country’s fiscal burden.

6. Conditions for connecting a common purchasing-power unit to a domestic budget

A unit used for international comparison and a currency paid out by a treasury serve different purposes. The appropriate conclusion is not that conversion from a common unit to domestic currency is universally impossible. National amounts, compatible conversion factors, the price reference year and the relevant period can together support a consistent conversion. The problem is that a single aggregate ratio does not supply those inputs. Separating the feasibility of conversion from the availability of its inputs also avoids an unnecessarily pessimistic argument.

For illustration, define qᵢ as the domestic-currency amount corresponding to one common unit in a country. Given that country’s common-unit cost Cᵢ, its domestic-currency amount under the same conditions is qᵢCᵢ. Multiplying the combined C for several countries by one market exchange rate, however, does not reconstruct each country’s domestic-currency amount qᵢCᵢ separately. Moving from a reference-year comparison unit to a current-year budget also requires accounting for subsequent price changes. Two numbers can both bear a dollar sign while serving different functions.

Matching estimates to domestic budget documents also requires checking the scope of expenditure. If a budget item includes business support and administrative expenses as well as transfers to households, a shared label does not make its coverage identical. Conversely, when household relief operates through lower tax revenue, looking only for a cash-transfer spending line will not reveal the corresponding whole. Translating a common-unit calculation into a national budget is a task of matching what the amounts count, as well as applying a conversion factor.

For businesses, this distinction directly affects estimates of government-supported demand. A large international estimate does not establish that public funds on that scale will be spent in a company’s sales territory. Conversely, an amount that looks small in local currency may still provide important support for the goods and services bought by the target households. Until comparable national evidence is available, decisions improve when the relevant products, households and expenditure definitions are matched, rather than turning an impression of scale directly into a demand forecast. Currency-related burdens are also discussed in Separate foreign prices from domestic-currency costs (Free guide).

7. Separate cost changes from changes in the denominator

A ratio can rise for reasons other than increasing compensation costs. Unchanged costs produce a higher ratio if GDP falls; costs can increase while the ratio declines if GDP grows faster. This basic decomposition is a way to identify causes, not to dismiss fiscal burdens. Without separating changes in programme design from changes in economic size, there is insufficient basis to call an improved ratio a policy achievement or a worsening ratio a programme failure.

In an international aggregate, denominator changes also alter country weights. Even if a country’s ratio is unchanged, its influence on the aggregate changes when its share of total GDP changes. Comparisons over time therefore require the same countries and a consistent price basis. An aggregate with newly added countries differs from a time series for a fixed set of countries. Treating missing countries as having zero burden would similarly turn a lack of information into an apparent improvement in the economy.

A mean should also be distinguished from a median. A GDP-weighted mean expresses the relationship between total cost and total GDP, but does not necessarily describe a typical country. The median is the middle value in an ordered list of countries and answers a different question. Neither is inherently the superior statistic: their roles depend on whether the question concerns an international total or the distribution across countries. Where possible, use both alongside national distributions and explain the economic-size weights in the aggregate. Basic checks are covered in Read denominators, periods and revisions (Free guide).

This matters for bond-market interpretation too. Investments are claims on specific issuers, whose revenue, spending, assets and liabilities matter. A decline in an international aggregate ratio alone cannot establish that a particular government’s capacity to pay has improved. The aggregate can nevertheless inform an assessment of how a shared price environment affects costs. Using it as background, while testing issuer-specific conclusions with national evidence, preserves useful information without overgeneralising it.

8. How country-level information changes the assessment

Pointing out an aggregate’s limitations is not enough to advance a decision. Additional evidence should be valued for the judgments it can change. National costs reveal where burdens fall; costs by income group show how coverage is distributed within countries. Neither replaces the other. Even with the same national total, costs concentrated in a few large losses can imply a different composition of consumption from costs spread across many small losses.

In a hypothetical international support allocation, the amount a country can finance itself matters alongside its needs. The country with the largest need does not necessarily require the most external support. But a shortage of external funding should not be used to reduce the estimate of need itself. Separating the calculation of household shortfalls from the calculation of who finances them helps keep three issues distinct: the need for support, domestic funding and external contributions.

Evidence for a fixed comparison group can reveal concentrations hidden by the aggregate. Comparing the cost of the same objective with each country’s revenue, for example, answers a different question from a GDP ratio. Yet mixing revenue definitions or fiscal years would make that new ratio incomparable too. Changing the denominator does not itself solve the problem: what the new denominator represents must be explained. Economic size, resources that can be collected and household losses are different quantities.

To make this process testable, it helps to specify in advance how new evidence will affect the judgment. If national ratios have little dispersion and conversion conditions are consistent, provisional decisions based on an average may be practical. If high burden ratios concentrate in small economies, an approach focused only on the total needs revision. Either result can be explained in terms of what previously missing distributional information revealed, rather than retrospectively praising or dismissing the average to suit the outcome.

9. Distinguish the stock below a poverty line from new downward crossings

To examine household conditions, introduce a separate illustrative assumption. Observe the same households at two dates, let y measure purchasing power per person per day on the same price basis, and fix a threshold L. Define date 0 as before the price change and date 1 as after it but before transfers. The report leaves the eligibility timing unspecified. The sets below use this explicitly shared assumption to investigate the difference between states and changes. Neither amounts nor counts are assigned to an actual country.

Households below the line at date 1 can have arrived there through two paths: some were already below it at date 0, while others were at or above it before moving below. Counting only the latter cannot establish the total at date 1. If households also cross in the opposite direction, the change in the total equals downward crossings minus upward crossings. Where population entries and exits are relevant, those must also be reconciled. The starting point is not to treat a stock and an inflow as the same thing.

This distinction remains necessary when a headcount improves. If households just below the line move above it while those far below deteriorate further, a better headcount does not establish that the most disadvantaged have been helped. Conversely, substantial improvement among households far below the line leaves the count unchanged if they do not cross it. A headcount is useful: it measures which side of a line people occupy. Asking the same indicator to measure their distance from the line goes beyond its function.

The meaning of the sets must also be preserved when several thresholds are compared. At a single date, people below a lower line are included among those below a higher one. But the sets of people newly crossing each line between two dates need not be nested in that way. Someone may cross the higher line without reaching the lower one; someone already below the higher line may cross the lower line. Before adding increases across thresholds, check whether the same people are included. The unit of aggregation—households or people—must also be fixed. Equal household counts need not mean equal populations when household sizes differ, and average household income cannot directly be compared with per-person income. Converting household-based records into people requires the corresponding household sizes. Omitting the unit risks falsely equating a population count with the number of benefit cases.

10. Equal changes in headcounts can conceal different depths of loss

Consider two hypothetical regions with the same number of people crossing below a line. In one, households move from just above to just below it; in the other, they fall from a higher starting point to far below. Headcounts cannot distinguish these situations, but the losses of purchasing power differ. Deriving equal compensation amounts from equal headcounts therefore requires further conditions on each household’s starting and ending positions. The counterexample does not discount headcounts: it identifies the information missing when moving from counts to amounts.

The treatment of time also changes the result. An equal shortfall in the same month has a different cumulative cost depending on whether it occurs once or continues for several periods. Relating a headcount at one point in a year to an annual cost requires defining the period connecting them. For households with seasonal income, simply switching observation months should not be interpreted as lasting improvement. More observations alone are insufficient; definitions of income or consumption per period must also align. The distinction between income and cash on hand is explained in Separate household income from cash needs (Free guide).

Purchasing baskets matter too. Even at the same total expenditure, the amount required to preserve earlier living conditions can differ with the share of products whose prices changed. An independent test separates product quantities and prices as well as comparing total spending. Nominal expenditure can rise while quantities fall because of higher prices; alternatively, switching products can leave total spending unchanged. The amount a household pays and what it can actually consume should not be treated as the same measure of recovery.

For business planning, this distinction affects regional demand assessments. Even if the same number of households deteriorates, sales follow different paths depending on whether the adjustment reduces essential purchases or postpones durable goods. Rather than mechanically multiplying headcounts into market size, examining the target households’ purchasing baskets and quantity changes makes business effects more specific. Spending changes among households outside the estimate’s coverage also need separate attention when explaining a company’s entire customer base.

11. The amount needed to regain the line versus earlier purchasing power

Under the preceding assumptions, let a household newly below the line satisfy y₀≥L>y₁. Its loss relative to earlier purchasing power is D=y₀−y₁, while the gap to the line is K=L−y₁. Therefore D−K=y₀−L. Unless the starting point was the line itself, reaching the line still leaves a gap to the earlier position. This is arithmetic between two destinations defined on the same scale, not an estimate from observations. The illustration also assumes that a transfer translates one for one into purchasing power.

The two amounts coincide under these assumptions when the starting point is exactly on the line. The higher above the line a household started, the larger the difference between returning to the line and returning to its previous position. This does not establish that particular households should receive priority. It means that treating restoration of the headcount and restoration of purchasing power as the same achievement requires additional conditions. Matching the outcome measured to the objective reduces the risk of either overstating or understating the value of the same expenditure.

Applying the reasoning to a household already below the line raises a different question. Returning to its position before the price change can still leave it below the line. Restoring a previous state and attaining a living standard at or above the threshold are different achievements. Calling the improvement ineffective because the headcount did not fall would confuse objectives. Yet concluding that a return to the earlier position guarantees an adequate living standard would also overlook the relationship with the selected threshold.

In real policy evaluation, the assumed one-for-one relationship between a transfer and purchasing power is itself testable. Simultaneous changes in prices, available products or household composition can produce different endpoints for equal transfers. That does not make the arithmetic distinction dispensable. Defining the desired state first makes it possible to investigate why the amount received and the change in living conditions failed to correspond. If the objective remains vague, it becomes harder to distinguish why costs rose from why effects were weak.

12. Keep thresholds and measurement dates consistent

In June 2025, the World Bank updated its main international poverty lines to 2021 purchasing power parity (PPP). The category using $6.85 under 2017 PPP now uses $8.30.[2] Changing only the threshold number does not turn the same counts or costs into results under the new basis. Updating comparisons requires checking income or consumption measurement, price conversion and threshold setting together.

An independent evaluation should also fix the date at which its target group is defined. If it uses the pre-transfer set assumed in section 9, a household subsequently lifted above the line by a transfer should not be retrospectively removed from that initial group. Mixing eligibility assessment and outcome assessment at the same date creates a circular evaluation in which the most successful recipients are treated as never having belonged to the target group. This is an analytical procedure for fixing a comparison population, not a determination of actual programme eligibility.

A comparison with a fixed starting cohort and a programme that continually updates eligibility can serve different purposes. A programme focused on people in continuing difficulty has reason to reflect changing circumstances. But data used to evaluate it must track new entrants and people leaving because their situation improved; otherwise lower expenditure cannot simply be read as lower need. It is important to explain separately the programme’s eligibility rule and the population definition used to measure its effects.

Similarly, comparing average incomes from different surveys may not show that the same households improved. Check whether samples or household composition changed, and distinguish evidence following the same people from representative snapshots at each date. When complete tracking is unavailable, aggregate comparisons can still be useful if that limitation is explicit. The requirement is not to pretend missing information has been supplied, but to identify which conclusions depend on the observation method.

13. When can lower cost be called a saving for the same outcome?

Calling the cheaper option more efficient requires comparable outcomes. For example, if programmes aim at the same endpoint for a fixed group, handle shortfalls consistently and one costs less, improved efficiency is an explanation worth examining. If coverage or the endpoint was narrowed to reduce cost, the first explanation is that the scope of the problem addressed has changed. This does not deny the reduction in cost. It classifies correctly what produced it.

An independent comparison can hold four conditions fixed: population, endpoint, observation period and cost coverage. When one differs, isolate that difference. Separating a change in population alone from a change in objective alone reduces the risk of attributing the entire expenditure gap to better programme operation. But invented numbers that assume these conditions are equal cannot be presented as empirical findings. The strength of an evaluation depends on whether corresponding evidence actually exists.

Costs also differ according to whether they measure transfers reaching households, amounts borne by government or resources used by society as a whole. If one entity’s spending falls because another takes over the burden, the first entity’s budget saving is not equivalent to a reduction in total cost. For example, requiring households to pay for the same service themselves can reduce government spending while moving them further from a living-standard recovery objective. Explaining what became more efficient requires tracking both the payer and the beneficiary.

This perspective does not settle the case for or against support in one direction. Narrow coverage may have a reasonable justification, and broad coverage may have one too. Making the difference explicit allows a smaller-budget solution to a narrower problem to be discussed alongside a measure protecting a wider set of conditions. Rather than compressing the dispute into the single word “efficiency”, comparisons that disclose which losses are being left unresolved make policy choices more accountable.

14. Connect national amounts and household objectives in one conditions table

Linking a national budget to household endpoints requires checking that separate documents refer to the same population. If the national cost used in an international comparison does not match domestically defined target households, an appropriate conversion factor still will not make the final amounts correspond. Checking population, monetary unit, price date and cost coverage in sequence reveals precisely what information is missing. Connecting two statistics in one calculation requires correspondence in country, population, unit and date.

Suppose, for example, that a national total includes all income losses while a domestic figure covers only particular products. Their difference may reflect coverage before it reflects conversion error or implementation failure. Analysts need either to compare only the corresponding portion or obtain evidence that allows both figures to be rebuilt on the same scope. Rather than immediately treating a discrepancy as a programme problem, starting with what each figure counts helps identify its cause.

The same conditions can guide revisions over time. When the price basis changes, record which of national cost, GDP and the conversion factor has been updated. Keeping partly updated figures separate from figures fully rebased to a common standard reduces the risk of mistaking methodological changes for changes in the economy. Comparability is better preserved by selecting mutually consistent inputs than by selecting each figure solely because it is newer.

Once the inputs are available, uncertainty about calculation can be separated from uncertainty about policy. Even when a national-currency amount can be calculated clearly, financing it and prioritising it against other expenditure remain separate choices. Conversely, a clear policy objective does not determine the amount if national cost evidence is absent. Moving from international discussion to a concrete domestic decision requires not treating agreement on one question as proof that the evidence for the other is available.

15. How household recovery reaches business demand around the world

Differences in the objective used to measure household conditions also affect forecasts of demand across borders. Even if more households receive support, the effect on domestic firms and import businesses depends on what those households buy. In our analysis, we hold the composition of purchases by the target households constant and distinguish whether recovered income buys additional quantities or covers higher prices for the same quantities. Reading growth in nominal sales alone as an expansion in demand volumes can lead to mistaken production and transport plans that depend on quantities.

If spending on imported goods is assumed to rise, a channel from domestic household support to overseas producers’ sales can be examined. Its size, however, depends on domestic distribution costs, the mix of imported and locally produced goods, and inventory adjustments. An additional unit reaching a household does not translate into an equal amount flowing abroad. Converting national costs into currency and measuring which country’s value added receives that spending are two tasks requiring different evidence.

In international resource markets, too, nominal compensation for higher prices should be distinguished from compensation that raises actual purchase volumes. If volumes can be maintained, compensation may have supported demand that would otherwise have fallen. Even without year-on-year growth in quantities, an effect can exist relative to the counterfactual. Establishing that effect requires a comparison that aligns prices and supply conditions other than compensation. A large spending amount alone does not reveal the direction or size of its effect on prices across the market.

When businesses use this analysis, they should select evidence for the household groups and regions corresponding to their own customers, rather than align their sales forecast geography with an international aggregate. Maintaining quantities by switching to cheaper products and reducing quantities of the same products have different implications for product mix and distribution. Assessing price competition, procurement and transport capacity requires looking beyond total compensation to spending allocations after support. Differences in household endpoints reach business analysis through these allocations and quantities.

16. Considering the argument that aggregates alone can support decisions

Aggregate information helps establish the scale of a problem while individual data are still being gathered. Refusing to use international comparisons because national conditions are not perfectly aligned can obscure common pressures. Thus, saying that national amounts cannot be uniquely recovered is not a conclusion that aggregates are worthless. The question is what precision the intended decision requires. Understanding the overall order of magnitude and setting a particular country’s budget envelope require different information.

When using an average as a practical approximation, one approach is to specify in advance the error boundary that would change the decision. For example, evidence can narrow the range of possible national costs, allowing an assessment of whether every value within that range leads to the same budget choice. If the choice is unchanged, further refinement can be deferred; if it changes, obtaining national inputs becomes a priority. The acceptable error should be defined in relation to the decision boundary, rather than selected afterwards because it is easy to calculate.

Another objection is that dividing objectives too finely complicates discussion and makes broad political agreement harder. Yet if the same language continues to contain different endpoints, assessments of success can diverge after agreement is reached. Not every indicator needs equal weight from the outset: one main objective can be chosen while other conditions are tracked through supporting indicators. What matters is recognising at the planning stage the problems that can remain even when the main indicator improves.

Evidence addressing this objection need not consist only of more complex models. Simple tables of the distribution of national ratios, income changes within the same group, and movements around a reference line can show how far an approximation remains useful. Complexity itself should not be treated as quality; additional information that changes a decision deserves priority. Conversely, if those tables adequately support an approximation based on the aggregate, there is less need to foreground every detailed distinction on every occasion. Analysis should identify the range in which a simple explanation holds, rather than seek to reject simple explanations.

17. Cost and the priority given to protection are different questions

Even a sound cost calculation does not automatically determine which households and which shortfalls should have priority. A policy prioritising a lower headcount and one prioritising relief of the deepest shortfalls can favour different recipients under the same funding constraint. Choosing between them involves social value judgments. Calculations can make those judgments more transparent, but a smaller cost does not make the unselected problem less important.

For an illustrative comparison, we separate the axis of keeping the target population fixed or expanding it from the axis of closing the distance to the line or the gap from the past. These combinations do not mean that all such programmes actually exist. They are a tool for checking whether the population and endpoint are being changed simultaneously when an objective is described. If one proposal is cheaper than another, the policy explanation differs depending on whether the reason lies in a changed population or a changed endpoint.

Even when supporting priority for households with the lowest incomes, that priority should not be presented as derived directly from a cost estimate. It needs a separate rationale, such as an objective of securing a minimum standard of living. Considering losses among households that remain above the line can also have reasons, including preventing a sudden fall in consumption or greater instability in living conditions. Stating each rationale makes clear where conclusions can diverge despite using the same evidence.

For market analysis, it is important not to erase these differences in value judgments by collapsing everything back into a single total. Even with the same budget, different purchase baskets among recipient households change the transmission to business demand. Conversely, different budgets may produce similar spending on products of interest to a business. Policy desirability and benefits to a particular company should not be judged on the same scale. The analysis should distinguish whose conditions improve from the demand changes that follow.

18. How decisions change across three states of information

The first information state is one in which national costs and conversion conditions are available and endpoints can also be observed for a fixed group of households. Discussion can then move from aggregates to comparisons of how much is needed in each country for each objective. An amount that becomes calculable is still different from a budget that has actually been approved. Greater calculation accuracy should not be substituted for a government decision to spend that amount. When assessing market expectations, distinguish whether the new information concerns needs, a policy proposal, a decision or an actual result.

The second state is one in which the total is clear but its national allocation is not sufficiently known. Discussion of the worldwide scale of the problem remains valuable, but assessments of a particular country depend on assumptions with a range of possible values. Instead of attaching an arbitrary numerical range, the analysis can specify which input the conclusion awaits. The next evidence to seek differs depending on whether a national ratio or a correspondence between cost categories is needed. This makes the decidable and undecidable parts clearer than summarising every gap with the single word ‘uncertainty’.

The third state is one in which the budget amount is known but household endpoints are not. Even if spending follows the plan, performance against the objective is unresolved without measuring changes in purchasing capacity. Information that more people are above the line still requires separate evidence on how far earlier consumption has recovered. Rather than price in the use of a budget as complete demand recovery, observations of quantities and spending composition are needed.

These states are not three scenarios forecasting a stronger or weaker crisis. They distinguish how specific a judgment can become with the available information. More information does not always improve the outlook: previously hidden national burdens or deep income losses may lead to a harsher assessment. Analysis should be updated according to which unknowns have been resolved and which assumptions no longer hold, rather than simply changing conclusions in the order that new numbers arrive.

19. Decide in advance which evidence would change the assessment

The evidence that would weaken this article’s central reservation about national budgets is clear. If compatible national costs, GDP, conversion factors and price bases are disclosed and allow domestic-currency amounts to be reproduced, many problems that aggregates alone could not resolve disappear. In that case, the explanation that national amounts are unknown should not continue.

For household endpoints, too, evidence that recovery to the line and recovery of earlier purchasing capacity are close in practice could reduce the practical importance of distinguishing them. Establishing that requires observing the starting points and endpoints of the same subjects. Similar averages for different groups do not prove that the same households have recovered to a similar extent. It matters which evidence connects a mathematical possibility to concrete circumstances.

Evidence that can challenge the proposed transmission to business demand includes purchases by product among the target households and the geographical distribution of their spending. If compensation increases but quantities of the product of interest do not move, an outlook assuming a strong demand effect for that product needs revision. Conversely, smaller quantity declines than in a comparison group may reveal support that a year-on-year comparison alone missed. If the conclusion changes with the choice of comparison group, that sensitivity should also be reported as part of the result.

When checking the consistency of evidence, examine whether subsequently revised definitions have been applied to past values. A succession of figures under the same heading is not necessarily a continuous series if coverage or the price basis changes. When revised evidence no longer supports the original interpretation, explaining which correspondence changed is more useful to readers than maintaining the earlier conclusion. Recording assumptions, observations and assessments separately helps show that a changed conclusion is not arbitrary.

20. Reconnect the size of the budget with the extent of recovered living standards

An aggregate cost ratio helps establish a shared understanding of an international problem. Moving that ratio into national budgets requires recovering national distributions and conversion inputs. Assessing household recovery further requires checking the starting point and intended endpoint. Separating these two tasks makes it possible to explain concretely what the same number can and cannot mean.

Policy comparisons should ask more than whether an amount is large or small. They should distinguish a changed population, a changed endpoint, and achieving the same objective with fewer resources. Even after the amount needed in each country is known, decisions on priorities and financing remain. Conversely, even when society shares an objective, the evidence needed to calculate a corresponding amount can be missing. Agreement and measurement are both necessary; neither can replace the other.

Effects on the world economy then emerge through what households can buy, what businesses can sell and which costs governments bear. The number of people returning above a line does not determine the recovery of consumption as a whole, and an international total does not determine a particular country’s financing needs. Maintaining these distinctions connects household support figures coherently to both fiscal conditions and business demand. It also clarifies which outlook should change as more individual evidence becomes available.

SG Group emphasises the inputs that identify national amounts and the evidence measuring endpoints for the same households. With these in place, analysis can move from identifying the limits of aggregation to comparing actual burdens and effects. Where they are missing, conclusions should remain conditional instead of filling the gaps with guesses. Assessing recovery in living conditions requires looking beyond a low cost to which shortfalls have been resolved and which remain.

References

1. No Time to Recover — UNDP. Cover: 12 October 2026. Release page.
2. Global Poverty Lines Update — World Bank, 5 June 2025.

The equations and hypothetical household classifications are independent analysis illustrating conditions for comparison. They do not identify a particular country’s budget, benefit eligibility or individual investment outcomes.