Why This
Matters
The proposed Family Card program could
represent the largest shift in social policy in Bangladesh’s history. At Tk
2,500 per month for around 40 million households, the annual cost would be
approximately Tk 1.2 trillion—close to 2 percent of GDP.
That scale is comparable to the entire current
social protection budget and similar to the national education allocation. It
exceeds public health spending.
For context, large targeted cash-transfer
programs in comparable countries typically cost around half of one percent of
GDP.
This is not simply a campaign pledge. It is a
structural commitment that, once introduced, becomes politically and socially
difficult to reverse—regardless of future economic conditions.
Before debating whether it is desirable, we
must address foundational questions:
Is the program universal or targeted?
How will it be financed—through taxation, borrowing, or reallocating existing
spending?
Will it consolidate existing programs (nearly 100 of them) or add another
layer?
What are the long-term fiscal implications if growth slows or inflation rises?
These are design questions. Design determines
sustainability.
The
Program’s Core Design
When a government proposes monthly transfers
to millions of families, the first question must be clear: Is it for
everyone—or only for those judged to be in need?
This distinction defines the program’s
structure, cost, administrative demands, and long-term durability.
Universal and targeted programs operate
differently. They create different expectations, require different systems, and
generate very different fiscal obligations.
The
Universal Path
A universal program means every household
qualifies. There is no income testing, no proxy means score, no screening
mechanism. The transfer is treated as a basic entitlement—closer to a
citizenship dividend than selective welfare.
Arguments in favor of universality include:
Administrative simplicity: In an economy dominated by informal employment, accurately identifying
the poor is complex and costly. Universality avoids that challenge.
No exclusion error: Poor households are not left out due to flawed data, outdated lists, or
discretionary interference.
Reduced stigma: No one must prove poverty to receive support. This matters for dignity
and social cohesion.
Political durability: When broad segments of society benefit, programs tend to be more
resilient to future budget cuts.
But universality has a clear fiscal cost. At
roughly Tk 1.2 trillion annually—around 2 percent of GDP—it requires explicit
trade-offs.
In a country where health, education,
infrastructure, and climate resilience remain underfunded, every taka allocated
here cannot be allocated elsewhere.
This is not a judgment. It is arithmetic.
The
Targeted Path
A targeted program restricts eligibility to
households identified as poor or vulnerable. The rationale is poverty
efficiency: Tk 2,500 has far greater impact for a household struggling to
afford basic necessities than for a middle-income family. Targeting
concentrates resources where welfare gains are highest.
It also reduces fiscal cost:
Covering the bottom 30 percent of households
would cost roughly 0.6 percent of GDP.
Covering the bottom 40 percent would cost approximately 0.8 percent of GDP.
However, targeting is administratively
demanding. It requires determining eligibility in an economy where incomes are
informal, seasonal, and difficult to verify. No targeting system eliminates
error:
Exclusion errors: Poor households left out due to imperfect data or flawed selection.
Inclusion errors: Non-poor households included, diluting resources meant
for the vulnerable.
These errors shape perceptions of fairness and
legitimacy.
There is also the “missing middle”
issue—households just above eligibility thresholds who may feel excluded while
not economically secure.
The Fiscal
Reality
Bangladesh’s tax-to-GDP ratio is approximately
9 percent and has dipped lower in recent estimates. This remains well below
many peer economies—roughly half that of India (17%) and significantly below
Indonesia (12%) or Vietnam (20%).
Fiscal space is therefore constrained.
With such a narrow revenue base, fiscal space
is structurally constrained.
A program of this scale must be financed
through some combination of:
Reallocation from other sectors: Cutting health, education, or infrastructure or other social safety net
programs.
Higher taxation: Expanding the tax base, which requires administrative
reform and political will.
Borrowing: Domestic or external, which carries debt sustainability and
inflation risks.
Sustainability is not a Year 1 question. It is
a Year 10 question.
If growth slows or borrowing costs rise, a
large structural transfer can crowd out other priorities or weaken
macroeconomic stability.
This is the arithmetic of choice.
The Unit
Problem: What Is a “Family”?
An additional design issue concerns the
definition of the beneficiary unit.
What counts as a “family”? A female head of
household? Or a unit linked to national ID?
In Bangladesh, family structures are
fluid—multi-generational living arrangements, temporary migration for work, and
shifts due to marriage or economic necessity.
If definitions are unclear, programs may
unintentionally create incentives for households to split or reclassify to
qualify for additional cards.
These behavioral responses are predictable.
They are rational adaptations to policy incentives.
Good policy anticipates them and builds clear,
transparent rules to prevent administrative disputes and unintended
distortions.
A
Foundational Question
We cannot meaningfully evaluate the Family
Card—its cost, fairness, or long-term feasibility—until we are clear whether it
is universal, targeted, or explicitly hybrid.
That clarity must come first.
In Part 2, I will examine targeting in
practice: how beneficiaries are identified, how accurate different methods are,
and why digital systems are not a magic solution to weak data.
The goal throughout this series is not
advocacy. It is informed reasoning.
Whatever design is chosen, it should be chosen
with eyes open—aware of the trade-offs, fiscal implications, and institutional
demands.
Tomorrow at 10am: Part 2 — Targeting in
Practice: How Do We Pick the Winners, and What Can Go Wrong?