And Then, Floods 1/

A critical macroeconomic assessment of IMF Conditionality on Kenya, 2021-present

Kwame Owino, Maureen Barasa, and Peter Doyle

July 15, 2024

Abstract

By April 2021, in the context of long-standing deep growth shortfalls, a heavily overvalued exchange
rate, an excessively loose scal stance, and an elevated public debt stock, Kenya’s prospects were
threatened by rising global interest rates, a large bullet payment due, and droughts.

What was needed was an IMF program to deliver an immediate change in the policy mix, with a sharp
front-loaded scal consolidation to allow a monetary loosening suf cient to correct the exchange rate
and inward orientation while keeping in ation on target. But the total scal correction should not have
been at the expense of medium-term growth, even if that required debt write-offs to reconcile it debt
sustainability. And the entire package should also have been resilience to further shocks.

But that was not the program that Kenya got. The program misdiagnosed misalignment, thus back-
loaded scal adjustment, and required medium-term primary scal balances well above global best
practice at the expense of growth potential, all re ected in relentless tax increases. And when its
conditionality on the Central Bank of Kenya turned out to be misspeci ed—including that in practice
it treated a large non-permanent relative food price shock as a matter only of in ation—that was not
corrected. So the program also delivered a monetary stance which was too tight, impeding the
necessary correction in the exchange rate, all at the expense of short-run growth as well.

These basic failures of quality control at the IMF meant that the program achieved neither its stated
goals nor the fundamental correction that was required—hence major nationwide social unrest.

A reset of the program for 2024/25 should be led by an immediate big relaxation in monetary policy, an
unchanged underlying primary balance outturn of a de cit of 1 percent of GDP remaining there
thereafter, leaving revenue ratio targets to the authorities, and activating a targeted program of income
support given food price shocks. If that requires debt write offs to secure sustainability, those should
be calibrated against a medium-term primary de cit of 1 percent of GDP. Alongside a major
retrenchment in the number of conditions and an increase in IMF transparency are necessary.
__________

1/ IEA Kenya acknowledges with thanks funding from IDEAs towards its contribution to this paper, and the
authors are most grateful to Jared Osoro, Gabriel Sterne, Germano Mwabu, Charles Abugre and Ndongo
Samba Sylla for incisive comments on an earlier draft. The usual disclaimers apply.
fi
fi
fi
fl
fi
fl
fi
fi
fi
fi
fi
fl
 Table of Contents

I. Ten Antecedents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . page 4

1. Low Growth
2. Misspeci ed Fiscal Policy
3. Real Effective Exchange Rate
4. Economic Orientation
5. Public Debt
6. Land
7. And then, Covid
8. And then, Global Interest Rates
9. And then, Drought
10. Stunting

II. IMF Program Framing and Priorities . . . . . . . . . . . . . . . . . . . . . . . . . page 19

1. Program Framing
2. Misalignment and Front-Loading
3. Medium-Term Primary Balance Targets
4. Public Debt Outlook
5. Fiscal Revenue Targets
6. Monetary Conditionality

III. Outturns Relative to Program Projections . . . . . . . . . . . . . . . . . . . . . . page 28

1. Deviations
2. In ation and Relative Prices
3. International Food Price Transmission
4. Warming
5. And then, Floods
6. Doom Loop
7. Christmas Tree Conditionality
8. Market Yield Spreads
9. Macroeconomic Judgement for 2024/25

IV. Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . page 46

Table

1. Deviations of Outturns From Program Projections, in percent, 2021-2024

Figures

1. Real GDP Per Capita, 2017 US International Dollars, 1990-2019

2. Primary Fiscal Balance, 1990-2021

2
fl
fi
 Table of Contents, cont’d

3. IMF Program Disbursements, 1984-2020

4. CPI Real Effective Exchange Rate of the Kenya Shilling, 2000-2023
5. Central Bank Policy Rates, 2000-2019
6. External Current Account Balance, in percent of GDP, 2000-2019
7. Export and Import Volumes of Goods and Services, Relative to Real GDP. 2000-2019
8. Private and Public Investment in GDP, and Real Machinery Imports, 2000-2019
9. Elasticity of Employment to GDP, 2000-2019
10. Public Debt, in Percent of GDP, 2000-2019
11. Real Output Across Covid
12. 10 Year US Treasury Bond Rate
13. Yield differentials, 10 year Government Bonds, Kenya less US $ Treasuries 2018-20
14. IMF Primary Balance Targets, in WEO Vintages
15. IMF Projections for Public Debt, in WEO Vintages
16. Revenue Ratios in Fastest Growing Countries, 1990-2019
17. IMF Recommended Revenue Ratios, in WEO Vintages
18. Central Bank of Kenya Policy Rate, 2019-2024
19. In ation, 2019-2024
20. Relative Price of Food, 2019-2024
21. Annual Real Wages De ated by Non-Food Prices, 2020-23
22. Consumer Price Indices by Income Group in Nairobi, 2019-23
23. Real Monthly Minimum Wages, Cities, in 2019 Low Income Shillings, 2019-2023
24. Daily Exchange Rate, Kenya Shillings per US$, 2019-2024
25. Non-Food In ation, 2019-24
26. Number of Formal Conditions at Origination and at Each Program Review
27. Yield Differentials, 10 Year Government Bonds, Kenya less US$, 2021-2024

3
fl
fl
fl
Since April 2021, the boundaries of Kenyan macroeconomic policy—within which policymakers have
discretion with IMF support—have been set by the IMF Extended Extended Fund Facility and
Extended Credit Facilities, supplemented since 2023 by its Resilience and Sustainability Facility.

Have those boundaries been well set?

I. Ten Antecedents

1. Low Growth

From 1990-2019, Kenya per capita real GDP, measured in 2017 International US$, grew by just a
quarter. That compares with the average of the eight highest real per capita GDP growth countries in
Kenya’s GDP per capita neighborhood which nearly quadrupled (Figure 1).

Figure 1. Real GDP Per Capita, 2017 US International Dollars, 1990-2019

Had per capita GDP in Kenya over that period grown at the same rate throughout as that of its best
performing peers, Kenyan real GDP per capita in 2019 would have been three times what it actually
was. That enormous per capita growth shortfall is apparent in all three of the sub-decades, so that most
of Kenya’s total growth over that period came from that of its population, not from its productivity.

4
 So in the per capita marathon, Kenya, exceptionally for any marathon and even after a head start, has
long been well off the leaders’ pace.

2. Misspeci ed Fiscal Policy

A misspeci ed scal stance is core to Kenya’s long-run per capita growth shortcomings—far too tight
from 1993-2010, then far too loose in the 2010s, with growth damaged by both errors.

Thus, the countries globally in Kenya’s GDP per capita neighborhood which achieved the fastest
growth of real GDP per capita 1990-2019 recorded average primary de cits (revenue less non-interest
spending) of some 1 percent of GDP. The interquartile range of the top eight of them was from
balance to minus 2 percent of GDP. And these countries maintained their primary balances inside this
interquartile range—the best peer band—for the bulk of this period.

Kenya’s record, measured against this best peer band, is shown in Figure 2.

Figure 2. Primary Fiscal Balance, 1990-2021 2/

__________

2/ Kenyan scal data to calendar year 2024 are all on a cash basis.

5
fi
fi
fi
fi
fi
Thus, rather than the prevailing narrative of of cial corruption and low quality spending—madharau—
from the mid-1990s to 2010, Kenya constrained the supply of public goods necessary for development
and incurred the X-inef ciencies from excess taxation so as to run primary balances far in excess of
optima for growth. It did so for the decade and a half to 2010, especially in the mid-1990s.

Those over-tight scal and associated lower-than-best-peer-growth outturns re ected quasi-

continuous IMF program conditionality which also addressed key matters in the composition of
revenue and spending, as well as structural policy matters (Figure 3).

Figure 3. IMF Program Disbursements, 1984-2020

As top marathon runners af rm, dietary discipline is essential. But no matter how well-speci ed the
rest of the exercise regime, starvation goes too far.

That privation reversed in the 2010s. Liberated from earlier IMF stringency by access to private
capital markets in an era the IMF applauded as “Africa rising”, Kenya then loosened its scal stance
considerably beyond the optima indicated by its best-performing peers.

The evidence from those best peers (and top marathoners) is that starvation is not corrected by
subsequent gluttony—but rather that long run growth is harmed by both errors, which compound.

6
fi
fi
fi
fi
fl
fi
fi
 In that light, several Kenyan economists in the 2010s correctly warned of the dangers of scal excess.
But these were “one-handed” economists as none of those condemned the earlier excess stringency.

The IMF, by contrast, was “two-handed” but in the wrong way—insisting on the earlier excess
stringency, then also welcoming the excess indulgence, wrongly on both counts.

Best-peer-matching scal policy throughout may not have suf ced for Kenya to match its best peers’
growth performance from 1990-2019. But the international evidence is that it was necessary for Kenya
to match their scal stance in order to have had any hope of matching their growth.

3. Real Effective Exchange Rate

A—if not the—key way in which the scal excesses of the 2010s transmitted into much lower-than-
best-peer growth outturns was by compromising Kenyan competitiveness.

The Kenya shilling appreciated some 40 percent between 2010 and 2019 (Figure 4).

Figure 4. CPI Real Effective Exchange Rate of the Kenya Shilling, 2000-2023

The misspeci ed scal stance during the 2010s is the apparent cause of this appreciation.

7
fi
fi
fi
fi
fi
fi
fi
 Following Dornbusch, in an open and small foreign exchange market with a oating exchange rate,
the scal expansions from 2010 onwards required the Central Bank of Kenya to raise its policy interest
rates to keep in ation within its target band of 5 percent plus a band of 2½ percentage points on either
side. That successfully kept average in ation at around 7½ percent annually both in the 2000s
excluding the Great Financial Crisis in 2008-09 and in the 2010s thereafter.

So in contrast to the early 1990s, when a brief destabilizing scal loosening in the context of a central
bank lacking independence manifested in a burst of in ation up to some 50 percent (Figure 2), the
strong policy response by the Central Bank of Kenya to scal loosening in the 2010s raised yield
differentials in Kenya’s favor to contain in ation, thus drawing capital into Kenya’s small forex market
to fund the de cit in a “risk on” phase in global capital markets, manifesting in Shilling overshooting.

Reforms in the monetary policy architecture complicate portrayal of these developments. Nevertheless,
it is clear that the differential between the Central Bank of Kenya policy rate and the Effective Federal
Funds Rate of some 3-5 percentage points in the 2000s widened in the 2010s to some 7-9 percentage
points (Figure 5).

Figure 5. Central Bank Policy Rates, 2000-2019

8
fi
fi
fl
fl
fl
fl
fi
fi
fl
Furthermore, the process of yield differential widening was serial. As “news” repeatedly came in
during the 2010s that the global “risk on” period was extending and that Kenya’s primary scal de cit
continued to deteriorate, so the corresponding serial unanticipated increases in the yield differential
elicited further capital in ows, and hence repeated episodes of Dornbusch overshooting, eventually
yielding a cumulative appreciation of a remarkable 40 percent in the CPI Real Effective Exchange
Rate over the decade.

This sequential/repeated Dornbusch overshooting account of the Kenyan shilling in the 2010s, rooted
in the analysis of global risk on and of Kenya’s primary balance relative to best-performing-peers,
offers a more compelling account of the appreciation than a “Balassa-Samuelson” account of it.
According to the latter, positive productivity shocks in the tradable sector drive up CPI real effective
exchange rates. But Kenya’s real GDP growth performance was lackluster in the relevant period
indicating absence of a particularly marked productivity shock (Figure 1). And if Balassa-Samuelson
was the driver, the external current account balance would not have deteriorated, let alone as sharply
and deeply as Kenya’s did (Figure 6).

Figure 6. External Current Account Balance, in percent of GDP, 2000-2019

And the Dornbusch-type explanation is also more compelling than a pure “capital in ows” account.
Exogenous surging capital in ows might explain an appreciating exchange rate and a rising current

9
fl
fl
fl
fi
fi
 account de cit but would do so alongside declines in yield spreads. But as noted above, those rose
markedly in the 2010s. And whereas capital surges usually exhibit for groups of countries, Tanzania
and Uganda did not exhibit similar exchange rate appreciations, but did avoid scal excess.

An examination of the political economy roots of the Dornbusch overshooting episode in Kenya after
2010—on both creditor and debtor sides—is beyond the scope of this paper. But its upshot was
macroeconomic. And congruent with that account of the exchange rate, the CPI-REER began to
depreciate in 2021 as scal retrenchment began (Figures 2 & 4).

4. Economic Orientation

The post 2010 marked appreciation of the Shilling was at the expense of Kenya’s outward orientation.

From 2000 to 2010, the volume of Kenyan exports rose by 40 percent relative to real GDP, a period of
export-led-growth (Figure 7). However, from 2011-19, alongside the marked appreciation of the
Kenyan Shilling, that totally reversed, so that by 2019, export volumes had grown only as much as real
GDP since 2000.

Figure 7. Export and Import Volumes of Goods and Services, Relative to Real GDP. 2000-2019

10
fi
fi
fi
 However, accompanying that remarkable gyration in the degree of export orientation of the economy,
import volumes broadly followed the same pattern, albeit from a lower peak and at a slower pace of
decline from 2010 than exports—even though the marked deterioration in competitiveness would
conventionally be expected to cause import intensity to rise rather than fall relative to GDP.

The explanation lies in the structure of Kenya’s imports. In 2010, over ⅓ of Kenya’s imports consisted
of industrial machinery—closely tied to total domestic, including public, xed investment.

In the absence of of cial data on volumes speci cally of industrial machinery imports, total industrial
machinery imports in US$, de ated by the US index for prices of machinery and equipment, after
rising strongly from the early 2000s fell sharply from its 2014 peak and then plateaued (blue line,
Figure 8). That broadly followed the share of total investment in GDP (red and green bars, Figure 8).

Figure 8. Private and Public Investment in GDP, and Real Machinery Imports, 2000-2019

Note that this estimation of the volumes of industrial machinery imports by Kenya using the US index
for prices of such goods invokes the law of one price for tradable goods. But it is quali ed to the extent
that the speci c product composition of such Kenyan imports may not fully match that of US
producers.

11
fi
fi
fl
fi
fi
fi
 The key insight here is that the elasticity of the largest single subset of Kenya’s imports—machinery
and equipment—with respect to the real effective exchange rate is negative. So in contrast to the
conventional sign which other import categories exhibit, overvaluation causes a large part of Kenya’s
imports, machinery, to fall, other things equal, because appreciation impairs Kenya’s standing as a
global production base, especially but not only for manufacturing.

That factor has to be taken into account when applying the standard techniques to assess currency
misalignment in Kenya. For Industrial Countries, for which those techniques and parameters were
developed and estimated, machinery looms large in both their exports and imports, so this particular
characteristic of machinery with respect to such means of measuring misalignment broadly cancel out.

But for Kenya—and others with highly structurally imbalanced shares of machinery between exports
and imports—a fundamental symptom of misalignment, low investment, reduces the large external
de cits which conventionally signal misalignment. This feature also distorts the trade elasticities and
trend exchange rate misalignment metrics of misalignment because it renders fundamental
overvaluation more sustainable in capital markets but at the expense of medium-term growth—which
is the ultimate metric of competitiveness.

Thus, all three conventional techniques used to measure misalignment estimated for and on industrial
countries systematically overstate competitiveness when applied to cases, like Kenya, where imports of
investment goods loom large in total imports and do not feature in exports. To date, the necessary
corrections to such metrics have not been made, including by the Central Bank of Kenya (2019).

Accordingly, the total export and import volume data (Figure 7) both con rm the deterioration in
Kenya’s outward orientation as the Shilling appreciated in the 2010s—with exports decelerating
markedly relative to GDP, and imports, dominated by machinery for investment, behaving likewise.

The negative impact on imports of machinery from the mid 2010s would have been even more evident
absent the strength public investment (green bars, Figure 8). Thus, the underlying demand for such
goods is indicated by private investment—which fell sharply from its 2007-13 norm in 2015-19 (red
bars, Figure 8), a decline that is even more precipitous for business (non-housing) investment.

That indication of business investment aversion re ected in the sharp decline in private investment
ratios after the 2007-13 norm suggests that the real effective exchange rate was broadly appropriate at
around the last time private investment was sustainably close to those norms, in 2011-13. This was
also the time at which export volumes began declining relative to GDP (Figure 7).

Given these indicators of private investment, machinery imports, and exports and their interpretation
—and with due regard to the error band of some 10 percentage points around any central point
estimate—the inference is that the Kenyan real effective exchange rate became overvalued from at
least 2013-2014—with the implication that by end 2019, it was overvalued by some 20 percent.

That estimate of misalignment for 2019 is far higher than others have suggested. But it is the outcome
fundamentally of a sustained weak scal stance in context of a small forex market, a risk-on global
environment, and an effective in ation- ghting Central Bank.

And it’s indication is rooted in private investors actual behavior from the mid 2010s, expressing stress
in declining private investment ratios and stagnant real machinery imports. That foundation of the
estimate of the degree of overvaluation is buttressed by declining export volumes relative to GDP, by

12
fi
fl
fi
fi
fl
fi
sustained below best-peer per capita growth, as well as by the standard metrics of misalignment when
corrected for Kenya’s particular import structure. And it is re ected in disappointing growth of
manufacturing, in a frequent preference for protectionist trade measures, including for foods, and in
the extent of the fall in private investment ratios, especially non-housing, in a global risk-on phase.

And its pattern is even commensurate with aggregate employment data over 2000-2019. The elasticity
of total employment to real GDP was highest and greater than unity—so employment grew faster than
GDP—in the 2000s when the real effective exchange rate was most competitive and the economy
strongly outward-oriented. But it declined in the 2010s when those conditions changed—falling to its
lowest level below unity at the end of the period when competitiveness was compromised (Figure 9).

Figure 9. Elasticity of Employment to GDP, 2000-2019

5. Public Debt

The overvaluation of the Shilling from the mid 20-teens attenuated the reported growth of public debt
by reducing the domestic valuation of foreign currency denominated debt.

Remarkably, none of the IMF program Debt Sustainability Analyses (DSAs) nor their stress tests nor
any of the Staff Report Risk Matrices consider the implications for the debt stock and IMF scal

13
fl
fi
policy counsel of the “contingency” that the comforting staff assessment of Shilling misalignment might
be wrong. This elementary professional lapse—to use an unduly generous term—includes that none of
those DSAs so much as report the foreign currency denominated share of the debt stock, let alone
draw any attention to it or run any illustrative simulations to assess its potential implications.

As ready reckoner, if all external debt at program eve is assumed foreign currency denominated and no
domestic debt is, the foreign currency share of public and publicly guaranteed debt was 2/3. If the
Shilling was 20 percent overvalued at that point, the headline debt stock to GDP ratio was
understated by some 8 percentage points of GDP. At the least, this shifts assessment of the strategic
options of adjusting to pay the debt versus write-offs in the direction of the latter.

Even absent this exchange rate correction, a key consequence of the radical shift in the scal stance
from 2010 (Figure 2) alongside weak growth was a signi cant rise in public debt ratios (Figure 10).

Figure 10. Public Debt, in Percent of GDP, 2000-2019 3/

_____

3/For the composition of Kenya’s public debt, at prevailing market exchange rates, by creditor, see
IMF Debt Sustainability Analysis p7, text table 2, here. Find the IMF presentation of the IMF/World
Bank Low Income Country Debt Sustainability Framework here.

14
fi
fi
 And when Kenya issued its US$ 2 billion bullet Eurobond in 2014, the settlement of which recently
proved so challenging, the IMF issued a full-throated endorsement of the credit, unsighted both to the
deterioration in competitiveness and associated harm to the economy’s outward orientation already
ongoing which the bond would aggravate, and to the impact on underlying public debt ratios.

So by end-2019, if debt is valued at the implied “correct” exchange rate—removing the estimated
overvaluation at that point—the public debt stock at end-2019 was close to 70 percent of GDP.

6. Land

That ratio is fruit not only of a half-decade of misalignment and three decades of second-best-primary
balances but also of six decades of misallocated land. That last is legacy of the colonialists who, by
Independence, had seized half the agricultural land, locking in that concentrated ownership thereafter
by insisting on full “willing-buyer-willing-seller” property rights as their last af iction bestowed on
Kenya before nally conceding democracy.

With less than 20 percent of Kenya comprising high-potential land where three-quarters of its people
live, that yielded extensive landlessness, squatting, congestion, and land clashes with half of the three
quarters of Kenyans who own land owning less than a hectare. And as, after full price sales, half of
arable land is now owned by just 20 percent of Kenyans, it has spawned a “big man” oligarchy, not
least as women comprise just 5 percent of registered land owners and the bulk of agricultural labor.

So, for most, the customary artery from land—via nourishment, employment, insurance, collateral,
livestock, marriage, and inheritance—to prosperity was severed, leaving many living short Malthusian
lives.

7. And then, Covid

So Kenya was highly vulnerable going into the Covid shock, even before that shock was ampli ed by
the Vaccine Apartheid practiced by the IMF’s dominant shareholders.

And so it was hit hard. But its limited scal resources meant that where Covid scal packages yielded
a change in the G7 primary scal balance in 2021 some 8 percentage points of GDP weaker than
anticipated in the Fall 2019 WEO, that difference was only 2½ percentage points in Kenya, due to the
relative modesty of its Covid scal package.

Underlying that package, with additional health spending, Kenyan Covid tax measures included full
income tax relief for monthly individual earnings below US$225; and cuts in the top pay-as-you-earn
rate from 30 to 25 percent, in the corporate income tax rate from 30 to 25 percent, in the turnover tax
rate on small businesses from 3 to 1 percent, and in the standard VAT rate from 16 to 14 percent,
partly offset by elimination of many exemptions in the VAT and corporate taxes. The consequent
revenue ratio fall was modest—less than ½ a percent of GDP—and temporary.

The Central Bank of Kenya did, however, follow other central banks cutting its policy rate, thus
avoiding what would otherwise have been yet further overvaluation of the already overvalued shilling.

And so the overall consequences for Kenyan real output were severe, with the accommodation and
restaurant sector particularly hard hit as tourism collapsed and the wholesale and retail sectors also
suffering as customers avoided stores (Figure 11).

15
fi
fi
fi
fi
fl
fi
fi
 Figure 11. Real Output Across Covid

But as it turned out, the global pandemic proved short-lived and by the time of the IMF program
design in Spring 2021, Kenya was already well on the way to full recovery to pre-Covid trend, even
though recovery for accommodation and restaurants took considerably longer.

8. And then, Global Interest Rates

Though the Covid hit to output turned out to be short-lived, so was the reprieve to Kenya’s
vulnerabilities from the sharp drop in global interest rates it occasioned.

By time of the IMF program design, long rates as indicated by the US 10 year treasury, were already
rising as markets began anticipating decisive in policy rate rises which then transpired (Figure 12).

16
 Figure 12. 10 Year US Treasury Bond Rate

For Kenya, the timing could hardly have been worse. After the full-some praise the IMF had given to
Kenya’s 10-year Eurobond on issue in 2014 and after the considerable deterioration in Kenya’s
competitiveness which it accommodated thereafter, its bullet 10-year maturity swung into view at the
precise moment that nancial markets began to close to emerging market borrowers.

9. And then, Droughts

Poor rainfall in ve seasons beginning in 2020 gave rise to the most severe drought in Kenyan records.
The challenges were ampli ed by Covid and by migratory desert locust infestations, greatly damaging
households as critical water shortages devastated food security and livelihoods, and conditions for
raising livestock, whether commercial or subsistence. When relief eventually came from these and
Covid, many (most?) households’ balance sheets—already fragile—had been signi cantly impaired.

10. Stunting

The shortcomings in macroeconomic policy over the past three decades in Kenya, often at IMF
insistence or indulgence, aggravated by recent shocks, are no intellectual abstraction. In 2021, as the
latest IMF program was being designed, one in ve of Kenya’s children were stunted, re ecting

17
fi
fi
fi
fi
fi
fl
severe, widespread, and enduring deprivation, even outside drought areas. Given widespread concern
with gender disparities, it is notable that this was concentrated disproportionately amongst boys.

On IMF program’s eve in 2021 therefore, all these ten antecedents were re ected in high yield spreads
between Kenya and global benchmarks, albeit off the recent mid-Covid highs, indicative of the
substantial and increased non-credibility of the prevailing policy framework (Figure 13).

Figure 13. Yield differentials, 10 year Government Bonds, Kenya less US$ Treasuries 2018-2021

18
fl
 II. IMF Program Framing and Priorities

By April 2021, Kenya needed an IMF program.

By then, already emerging from Covid, it had a long-standing deep growth shortfall, a heavily
overvalued exchange rate primarily re ecting an excessively loose scal stance, consequently a
dif cult—and understated—challenge to manage its public debt stock, an adverse context of rising
global interest rates for commercial re nance for that stock, a large bullet payment on it looming, and
ongoing droughts.

Absent an IMF program, all of that was set to make prospects for Kenya’s shockingly high proportion
of stunted children even worse.

What was needed was an IMF program to address those matters decisively, notably an immediate
change in the policy mix with a sharp front-loaded scal consolidation to allow a monetary loosening
suf cient to correct the exchange rate and inward orientation while keeping in ation on target.

But that total scal correction should not be excessive, at the expense of medium-term growth, even if
such a scal path required debt write-offs to reconcile it with a declining public debt to GDP ratio, the
conventional and appropriate overarching metric of debt sustainability.

In the context of the given framework, monetary conditionality needed to be adaptive. And the entire
program should have been designed to secure resilience to risk of further adverse exogenous shocks.

But that is not what happened.

Instead, the IMF back-loaded the scal adjustment while mandating medium-term primary scal
balances well above global best practice at the expense of medium-term growth potential, thus
subordinating all other goals to full debt repayment, never even mentioning the exchange rate
expressly or otherwise as a core object of macroeconomic policy.

And when its conditionality on the Central Bank of Kenya turned out to be misspeci ed in the
evolving context, that was not corrected. So it also delivered a monetary stance which was too tight,
impeding the necessary correction in the exchange rate, all at the expense of short-run growth as well.

In that context, the program’s stated goals have proved elusive, as has the fundamental correction in
economic orientation that was required. And the program’s consequent underlying fragility has been
most obviously re ected in the seismic ongoing social and political upheaval which it has now birthed.

1. Program Framing

Speci cally, the IMF headlined its framing in 2021 thus:

“The Fund-supported program will anchor the next phase of the authorities'
pandemic response and a multi-year consolidation effort centered on raising tax
revenues and rm expenditure control. By bringing the primary de cit below its debt-
stabilizing level during the program period, debt as a share of GDP would be put
rmly on a downward trajectory. Tax revenue would be brought back from current
low levels to the levels achieved in recent years … “

19
fi
fi
fi
fi
fi
fi
fi
fl
fi
fi
fl
fi
fi
fi
fl
fi
fi
 With the pandemic already largely receding, there is no mention here of growth, competitiveness,
economic orientation, drought, land, or stunting. Instead, again without justifying why these matters
were disregarded, the IMF framed the challenge purely as putting public debt on a downward
trajectory via adjustments to the primary balance, spearheaded by action on the revenue side, all
supported by what in Kenya are perennial commitments to reform State Owned Enterprises.

This was not a matter of prioritization or division of labor with other international institutions. The
most notable aspect of the IMF Staff Report which initiated the program is that, competitiveness
aside, it does not even raise these other matters, let alone explain how they should effect policy on debt
stabilization, nor prioritize among them nor, assign them to other institutions to address.

Instead, all too evident is a complete lack of curiosity to understand major prior macroeconomic
transitions in Kenya or their implications for program design—or perhaps it was absolute conviction in
the pre-conceived framing precluding any sense of need to verify it, or worst, a mindset unable to
conceive that Kenya (Africa?) might, technically, be macro-economically distinct and complex.

Whichever it is, after a brief reference to the IMF’s Covid SDDI debt service delay initiative—which
Kenya reluctantly joined even though most other eligible countries declined and despite creditors’
refusal to extend Kenyan payments falling due after mid 2021—the analysis reads like an elementary
spreadsheet exercise in the standard debt equation tying the debt trajectory to a given growth and
interest rate outlook, and primary balances.

Though IMF staff rightly emphasize risk of debt distress and the dif culties of attaining their primary
balance targets, there is not so much as a whisper about general debt service re-phasing—let alone
write offs—as an option to reconcile sustainability with growth, let alone any explanation of why such
policy options were ruled out or any indication if they were even considered.

This last emerges a serious recurring theme throughout—of core strategic decisions and numbers not
being justi ed by the IMF or, in many instances, discussed at all. Instead, they are implicitly presented
as axiomatic or as fait accomplis.

2. Misalignment and Front-Loading

A partial exception to this lack of explanation is competitiveness, where the External Sector
Assessment in the program initiating Staff Report uses the three standard metrics to assess
misalignment.

However, the discussion is boilerplate. It does not consider let alone counter the methodological issues
and evidence outlined above, nor evince any obligation to explain the 40 percent real appreciation over
the prior decade—dismissing that, without reference to the number, as a recent “trend” (as if those
don’t have to be interrogated)—nor the behavior of export and import volumes relative to GDP or the
large sustained external current account de cits alongside.

Instead, just “letting the three-component misalignment metrics sausage machine” spit Kenya out
without actually thinking about what it was doing, IMF staff conclude that in April 2021, the Shilling
was within 3 and 9.1 percent of its correct value. In so doing, they were oblivious to the absurd
precision in their upper bound given that in applied macroeconomics, even a conservative margin of
error is some +/- 5 percentage points around any central estimate of misalignment.

20
fi
fi
fi
But the core implication of the staff failure to identify the extent of misalignment was that the scal
adjustment path they recommended was too back-loaded. A sharp up front adjustment was needed to
allow monetary policy—via interest rate cuts—to drive a correction in the exchange rate. But not so.
Indeed, a proprietary study by Oxford Economics of some 200 episodes of scal adjustment nds that
the path for the Kenyan scal balance was exceptionally back-loaded.

And staff did not so much as consider in their Risk Assessment Matrix the implications misalignment
could have for core policy advice and debt sustainability if it turned out to be signi cant.

3. Medium-Term Primary Balance Target

Furthermore, by anchoring their scal targets in public debt reduction rather than in misalignment
and growth, the IMF also went far too far in its overall demands for scal adjustment.

In particular, the sole near constant in the program throughout is the IMF’s medium-term target for
Kenya’s primary balance, set to be close to a surplus of around 1 percent of GDP. That number
persisted despite multiple shocks since program initiation—including revised estimates of the 2021
debt stock, the outturn for the post-covid recovery in Kenya, and global interest rates rising higher
than anticipated.

However, neither the initiating staff report nor its successors explain this all-but-unchanging pivotal
target number—why not higher, why not lower? Furthermore, the texts do not even mention it (not
even, remarkably, the program originating Debt Sustainability Analysis) beyond offering generic
unnumbered exhortations about the necessity to put public debt “decisively” on a downward trajectory
with “ambitious” scal policy.

An effort to reverse-engineer the IMF’s rationale for that number by examining its evolution over time
yields little light. Figure 14 reports the IMF targets for Kenya’s primary balance across various
vintages of the WEO, starting with two immediately prior to the program. The further to the right the
lines extend, the more recent the WEO vintage. The initiating program is the red line, the most recent
vintage is the green line, and the outturns, including staff estimates for 2024, are shown in black.

21
fi
fi
fi
fi
fi
fi
fi
fi
 Figure 14. IMF Primary Balance Targets, in WEO Vintages

Immediately apparent is the dramatic change in the IMF targets at program initiation (red line)
compared to immediately prior it (dark purple). In just six months from October 2020 to April 2021,
the IMF raised its view of the necessary medium-term primary balance by 4 percentage points of GDP
—a sum equivalent to 1/4 of Kenya’s primary spending—without feeling any obligation to explain.

But following that extraordinary volte face, the original program targeted a medium-term primary
surplus of 1.1 percent of GDP (Figure 14, red line). That was revised moderately downwards in
subsequent reviews as summarized from each subsequent vintage of the IMF WEO, dropping to a
little below 1 percent of GDP (blue and yellow lines), and up again in Spring 2024 (green line).

However, where the pre program targets were well below best peer practice for countries in Kenya’s
GDP per capita neighborhood, the program medium-term target numbers are signi cantly above.

In particular, based on WEO worldwide IMF data from 1990-2019, countries which in that period
grew fastest at around Kenya’s GDP per capita averaged primary de cits of between 0.2 and 2 percent
of GDP (greyed box “Best Peer Target Band” in Figure 14). Relative to the center of that band, which
indicates a growth-optimal medium-term primary balance at Kenya’s level of GDP per capita of a

22
fi
fi
 de cit of 1 percent of GDP, the IMF target for Kenya is now some 2 percentage points of GDP above
global best practice having been well below that measure of best practice pre-program.

No rationale is provided or apparent for either of these deviations of core IMF targets for the primary
balance from global best practice, nor for why the IMF targets dramatically changed at program
initiation, and (so far) ever after.

Given long growth underperformance in Kenya—derived in signi cant part from earlier deviations of
primary balances from the target band (Figure 2)—the pre-program targets were set to continue that
underperformance by undershooting best practice, while those in the program go well beyond what
can be justi ed by such growth optimization.

4. Public Debt Outlook

Nor, even, can any of these targets be reverse engineered from the evolving outlook for public debt.

Even leaving aside that the public debt to GDP ratios are attered—and signi cantly—by
misalignment of the Shilling, the most notable feature of debt projections through each program
review, again as summarized by each vintage of the WEO in the same was as above, is that the
apparent anchor target of a primary surplus of a little below 1 percent of GDP was associated,
according to IMF projections, with wildly varying projections for public debt ratios (Figure 15).

Figure 15. IMF Projections for Public Debt, in WEO Vintages

23
fi
fi
fl
fi
fi
 Thus, whereas the pre Covid targets saw public debt stable at a little over 60 percent of GDP (light
purple line), the subsequent pre-program targets saw public debt rising sharply towards 80 percent of
GDP—but with the IMF lowering its schedule of required primary balance targets from its prior
vintage (Figure 15), formally indicating that it was unperturbed by that public debt outlook.

However, the program initiating staff report projected a debt ratio of 68 percent of GDP by 2026 (red
line)—thereby implicitly justifying the volte face on the necessary medium-term primary balance.

Just a year later, the IMF projected public debt in 2026 some 10 percentage points of GDP below that
(blue lines). Nevertheless, it left its primary balance targets broadly unchanged (Figure 14), without
explaining why having just accepted debt at 68, 58 percent of GDP in 2026 was suddenly necessary.

In Spring 2024, with the debt outturn for 2024 estimated above target for that year (black line), the
debt outlook for 2026 had reverted to the original program projection, with the medium term primary
balance target raised a little higher, and the trajectory to that medium-target raised very signi cantly
higher, and underlying sharp further tax rises in the 2024/25 budget (Figure 13, green line).

Given these enormous gyrations in the public debt to GDP projections with an unchanged primary
balance targets as the program evolved, it is evident that speci c debt ratio objectives do not underlie
the IMF’s calibration of the medium-term primary balance target of a little below 1 percent of GDP.
Had they been, the medium-term primary balance target would have been signi cantly adjusted in
2022 in light of the then-projected precipitous fall in public debt ratios, and it was not.

Accordingly, it is hard to discern any rhyme or reason from reverse engineering these medium-term
IMF program primary balance target numbers—they give every appearance of pure at, and chaotic
at at that.

And they require justi cation if for no other reason than that they are clearly above—and signi cantly
so—global best practice for growth and are bereft of evident connection to speci c debt trajectories.

5. Fiscal Revenue Targets

And those targets also essentially drive the speci c—and highly contentious—tax reforms to yield the
program revenue targets.

In stark contrast to its treatment of the pivotal medium-term primary balance target and its
relationship—if any—to the debt trajectory, the program initiating staff report was explicit and
detailed about the targets for revenue derived directly from them. With the Covid-stimulus revenue
measures already reversed before the program started, the aim was to reverse immediately all the
stimuli revenue losses. Thus:

“A key pillar of the strategy is to bring the tax-to-GDP ratio back to levels achieved
in recent years (from 12.9 percent of GDP in FY20/21 to 15.6 percent in FY23/24),
so as to generate resources to meet Kenya's development needs.“

Though speci c about motivation and numbers, the IMF case was nevertheless unpersuasive on both.

As much else, the key phrase “to meet Kenya’s development needs” is never elaborated or explained,
but is simply presumed to be self-evident, given the targets for the medium-term primary balances.

24
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
But it is not.

The fastest realGDP per capita growing countries globally from 1990 to 2019 secured that
achievement with a wide range of revenue to GDP ratios. Thus, for each level of GDP per capita, the
revenue ratios of the eight fastest per capita growing countries over that period can be ranked from
highest to lowest, and their interquartile range plotted.

These data are shown in Figure 16.

Figure 16. Revenue Ratios in Fastest Growing Countries, 1990-2019

The blue dots show the upper end of the revenue interquartile range, and the black dots the lower end,
with the red dots the average, all for the fastest real per capita growing countries 1990-2019.

The width of the interquartile range—indicated by the arrows, typically 20 percentage points—shows
that the choice of growth strategy by the most successful countries has varied widely. Some, at the
lower revenue end, have chosen to grow with limited impediments from taxation matched by similar
low provision of public goods for development. At the other end of the spectrum, other similarly
rapidly growing countries have achieved their success despite the impediments from high revenue
ratios by matching them with generous and effective provision of public goods for development.

25
Thus, there is no “single optimal ratio” of revenue to meet a country’s “development needs.”

So any revenue objective has to be justi ed relative to the country’s speci c and chosen development
strategy. And in Kenya, any case to raise revenue ratios must explain why the public’s reservations,
due to warranted concern with public waste and corruption, should be disregarded. The IMF does not
do any of this for Kenya, let alone explain why the revenue targets in its program are so aggressive.

It compounded these omissions in mid-program by comparing Kenya’s revenue ratio to that of other
African countries, replicating an exercise often conducted by analysts and reported by the Kenyan
authorities.

That comparator exercise misleads. The relevant peer group is not African countries—which not only
includes many with very different levels of GDP per capita from Kenya but also some highly sub-par
growth performers. Instead, the relevant peer group for this exercise (as for any aspirant marathon
champion) is to not to follow the example of a mix of middle-of-the-roaders, whether African or not,
but of the pack leaders, namely the globally fastest growing at Kenya’s level of GDP per capita.

On that basis, Kenyan revenue ratio for 2021 (green circle, Figure 16) is a little below the best peer
average, but close to the center of the interquartile range, thus showing no sign of being any sort of
obvious outlier. After the originally mandated revenue measures including for 2024, the IMF projected
Kenya’s revenue ratio to increase to the average of the interquartile range (brown dot, Figure 16).

That change may look modest. But tax adjustment on that scale and timing, as is now all-too-evident,
is the stuff of uprising—so its necessity must be fully elaborated. Yet no such explanation is supplied.

That does not mean the action is wrong, only that the necessity of those target numbers at that time
has not been laid out. Nor has their incorporation in IMF conditionality at all—as opposed to their
being left, along with expenditure totals, to the full discretion of the Kenyan authorities, both to be set
subject to the IMF’s primary balance targets.

In absence of those explanations, the inference is that the IMF has strongly mandated revenue ratio
increases in the program not, as claimed, in pursuit of a “development agenda”, but solely to shore up
(in its judgement) prospect of delivery of its misspeci ed medium-term primary balance targets. And
as those exceed the best growth optima, the inference is that the associated revenue targets, far from
supporting a “development agenda” are, by association with those primary balance targets, hostile to it.

Thus, after wildly varying trajectories in pre Covid WEO vintages, the initiating program raised IMF
targets for the revenue ratio sharply upwards by 3 percentage points of GDP (Figure 17, red line),
then subsequent vintages of the WEO revised them somewhat down, until the March 2024 WEO
revised them sharply upwards again (green line).

26
fi
fi
fi
 Figure 17. IMF Recommended Revenue Ratios, in WEO Vintages.

The practical consequences of all this are stark. Beyond headline tax policy changes—eg, abolition of
the reduced VAT rate for fuel and of the zero-rating of bread—the Kenyan Institute of Economic
Affairs records that the four Finance Bills from 2021 to the one for 2024 just rejected incorporated 66,
53, 84, and 69 changes to tax policy and regulations respectively.

Though the last of these was summarized prior to outbreak of mass protest by the President’s Chief
Economic Advisor as “we have to come through with a couple of things”, this fruit of aggressive IMF
revenue targets has even elicited a late caution from the IMF’s conjoined twin, the World Bank.

And it is not only generating chronic cumulative instability in the tax environment, but by being driven
under relentless external duress, is also undermining the proper checks and balances between the
executive and the legislature.

Given that the primary balance was on program target for 2023 (Figure 14), the only apparent reason
for the highly contentious and further very sharp increase in schedule of IMF revenue targets
mandated in Spring 2024 is that the public debt outturn in 2023 was higher than targeted and that the
IMF has, in light of outturns discussed below, marked down its medium-term growth outlook.

27
 6. Monetary Conditionality

In this program context, the IMF required the Central Bank of Kenya to “consult” on its
determination of its policy rate if headline in ation measured over 12 months deviated from its target
band—of 5 percent +/- 2½ percent for three consecutive months.

That particular arrangement, a speci c instance of the broadly standard means of reconciling
necessary IMF program monetary conditionality with the independence of Central Banks subject to it,
turned out to be seriously overtaken by events, notably by the Ukraine shock and droughts.

But no adjustments to the consultation arrangements were made in light of those events, leading
eventually, as discussed below, to signi cant misspeci cation of monetary policy, notably from 2024.

III. Outturns Relative to Program Projections

The overall critique of the design of IMF program conditionality is that scal adjustment was not
front-loaded to rebalance with monetary immediately to facilitate the required exchange rate
adjustment, but was instead backloaded, targeting excessive primary balance surpluses. That is
re ected in signi cant deviations of outturns from the original IMF projections under its policies.

1. Deviations

This comparison between projections and outturns can be made meaningfully because the Kenyan
authorities have gone to extraordinary lengths to comply with the IMF critical conditionalities—
however ill-designed. All program reviews have been completed with almost all IMF critical
conditionality—namely performance criteria and structural benchmarks—delivered on time and in
full. So unanticipated exogenous shocks aside, outturns re ect IMF diagnosis and policy mandates,
not failures of implementation.

This comparison of projections and outturns is summarized in Table 1, with outturns for 2024 taken
from Spring 2024 IMF WEO.

28
fl
fi
fi
fi
fl
fi
fl
fi
 Table 1. Deviations of Outturns From Program Projections, in percent, 2021-2024

Thus, as reported in the green shaded cells, in 2024, the level of Global real GDP is set on Spring 2024
IMF staff projections to be 1.6 percent lower than was projected at IMF program initiation.

In that context, Kenyan real GDP is set to be 2.9 percent lower than projected. That is a signi cant
shortfall—given that with real annual Kenyan growth typically around 5 percent, that is a projection
error in just three years of over half of one year’s typical growth. The IMF staff have not indicated that
long-covid health or psychiatric issues among the employed play any signi cant role in this. And the
trend underperformance of manufacturing and of wholesale and retail sales relative to GDP re ects
the underlying misalignment problem and household income stress respectively (Figure 11).

Alongside, the consumer price level in 2024 is set to be 8 percent higher than projected by the
originating program, in the context of exogenous shocks from Ukraine and droughts since 2022.

The increase in the revenue share in GDP is set to be 0.7 percentage points higher than programmed,
but the increase in the spending share is set to be 1.6 percentage points higher, the combination leading
to an overshoot of the scal de cit of 0.9 percentage points of GDP.

29
fi
fi
fi
fi
fl
 But that almost entirely re ects higher than programmed interest spending as the primary balance is
almost exactly on the target set in the initiating program,

In that light, public debt ratios are set in 2024 to have risen by 4.4 percentage points of GDP more
than initially programmed, but this is offset by the large downward revisions to the base in 2021 for
this indicator, so that the headline public debt to GDP ratio is just 1.2 percent of GDP higher than
originally anticipated.

Note, however, that with the scal squeeze in Kenya, the external current account balance is 1.6
percentage points of GDP stronger than originally anticipated, despite global output in 2024—and
hence demand—being 1.6 percent weaker than was projected. So the shortfall in Kenyan real output
relative to original projections was not because global output was weaker than projected. Despite
global weakness, the external current account was stronger than projected, thus supplying greater
than anticipated net external demand to Kenya.

So in 2021, the IMF considerably overstated Kenyan output potential and ability to carry its debt
burden—very much the original sin of the entire program—by overstating its growth prospects, by
misreading exchange rate misalignment and hence understating its public debt stock, and by imposing
medium-term primary balance targets which compromise its growth potential.

These errors largely re ected the IMF’s misdiagnosis of the Kenyan economy as discussed in
Section I, and were material given that they arose shortly after Zambia defaulted and just before
Ghana followed suit, both thereby triggering debt restructurings.

Nevertheless, these outturns relative to initiating program projections leave a question. If a scal
consolidation was indeed required to deliver a large exchange rate correction, why was the actual real
effective exchange rate correction so modest in 2022-23 and since largely reversed in 2024, when the
targeted scal consolidation was excessive (Figure 4)?

The answer is that the critique of the primary balance targets refers to their impact on medium-term
growth on the supply side, whereas the exchange rate, insofar as it responds to scal policy, primarily
re ects its short-to-medium-term demand side effects. And not only was the targeted scal trajectory
too backloaded but the outturn to 2024, measured by the change in the overall balance, was little less
than even the IMF had targeted (Table 1). The remainder of the answer lies, despite headline in ation
overshooting the target band for an extended period, in chronic over-tight monetary policy in the
context of Ukraine, drought, and default risk, which has now become severe.

2. In ation and Relative Prices

In concordance with IMF directives, the Central Bank of Kenya responded to the Ukraine shock,
market fears of a default on the maturing Eurobond, rising global interest rates, and drought from
2022 by raising its policy rate in a series of steps from 7 to 13 percent to address headline in ation
rising towards the top of the target band and then overshooting it (Figures 18 & 19).

30
fl
fi
fl
fl
fl
fi
fi
fi
fi
fl
fl
Figure 18. Central Bank of Kenya Policy Rate, 2019-2024

31
 Figure 19. In ation, 2019-2024

Conditions for monetary policymaking were dif cult. But that merely raises the premium on high
quality analysis, in the absence of which the policy responses were overdone.

That excess occurred despite IMF staff getting three core matters in their monetary conditionality
right: that there was no critical need for any reform of the Central Bank of Kenya’s In ation Targeting
monetary framework; that the monetary consultation arrangement with the IMF should be symmetric;
and that after supply shocks, monetary policy should be on the lookout for “second round” impacts on
in ation.

But the IMF consultation clause was wrongly de ned on a rolling three month average of the 12-
month rates of in ation rather than on rolling averages of in ation over three months seasonally-
adjusted annualized, thus all-but-guaranteeing that formal IMF interventions would be systematically
“behind the curve”. And following the supply shocks, even that was not rede ned from headline onto a
measure of underlying in ation.

These technical errors were compounded by the assessment that the exchange rate was not far adrift of
fair value rather than signi cantly misaligned. So the IMF interpreted Shilling weakening as a sign of
monetary stimulus and loss of investor con dence rather than as a necessary correction.

32
fl
fl
fl
fl
fi
fi
fi
fi
fl
fi
fl
 And the errors were further compounded by the overall IMF treatment of food price shocks as matters
of general in ation rather than as re ecting large enduring but non-permanent relative price shocks.
Indeed, IMF staff seem unaware—as they have not even reported the data, referring only to a “cost of
living crisis” as if that is synonymous with in ation—that after the Russian invasion of Ukraine and
local droughts, the relative price of food has risen by 30 percent (Figure 20).

Figure 20. Relative Price of Food, 2019-2024

Nor, because they do not report these data either, is there any sign that they conducted ongoing checks
on the critical “second round effects“ metric, namely in wages.

On that, broadly, with trend real GDP and employment around 5 and 3 percent respectively, there is
room for annual non-in ationary real wage growth of some 2 percent. However, average wages
de ated by non-food prices have fallen continuously since 2020 to 2023, in both the private and public
sectors Figure 21). Thus, there is no evidence of “second round effects” in data to 2023.

33
fl
fl
fl
fl
fl
 Figure 21. Annual Real Wages, Nominal Wages De ated by Non-Food Prices, 2020-23

Nor is there informal evidence thereafter—including from wage settlements data, income tax receipts,
or the Central Bank of Kenya’s market information gathering process. But nowhere is this re ected in
staff assessments. Despite their appropriate second-round warnings, the IMF staff simply do not
report, assess, or address nominal or real wage data, nor explain why not.

And most remarkably, given the centrality of wage data to detecting the presence or otherwise of
second round effects, the IMF staff report annexes assessing the quality of various key macroeconomic
statistical series do not even list the Kenyan wage data series, let alone assess their quality.

But with food weighted at 33 percent in the CPI basket, the enduring but analytically temporary food
shock took the biggest toll on the poorest for whom that share in spending is highest. This is evident in
the particular elevation of prices for low income groups in Nairobi (Figure 22).

34
fl
fl
 Figure 22. Consumer Price Indices by Income Group in Nairobi, 2019-23

A measure of the toll on households at the low end of the income spectrum may also be inferred from
city—Nairobi, Mombasa, Kisumu, and Nakuru—minimum wages, de ated by the Nairobi low income
consumer price index. After the prior increase in early 2018, these fell a further quarter from January
2019 to December 2023 (Figure 23). With low income prices still rising disproportionately, nominal
minimum wages were raised by just 6 percent in May 2024. All this further not only re ects the
absence of second round effects even after 2022, but rather it suggests the stark contrary.

35
fl
fl
 Figure 23. Real Monthly Minimum Wages, Cities, in 2019 Low Income Shillings, 2019-2023

In the context of major sustained relative price changes concentrated on the poor and of Malthusian
living by many due to the fundamental mis-allocation of land, standard virtue-signaling IMF phrasing
that “social spending” under the program is being maintained rings particularly hollow.

Given overall policy anchored by a determined delivery of a substantial (and in the medium term
excessive) scal consolidation at IMF insistence, the appropriate IMF response to the food price shock
would have been: to call for a temporary targeted scal program of income support for those on lowest
incomes facing the biggest temporary real income shock; to welcome the accompanying exchange rate
correction; and by re-specifying the monetary conditionality to focus not on the headline rate of
in ation over 12 months but on the underlying rate as indicated by non-food in ation over three
months, to direct the Central Bank of Kenya to “see through” the temporary supply shocks and make
full use of the exibility provided in its monetary framework in regard to its target horizon to do so.

But that is not what happened. Instead, the IMF accommodated no relief for the poorest, it kept the
headline CPI as the metric for its consultation requirement rather than re-specifying it to non-food
in ation, and accordingly it insisted on considerable monetary tightening in response to headline
in ation numbers despite the absence of evidence of second round effects. Thus, after a series of small

36
fl
fl
fl
fi
fl
fi
fl
steps and after the IMF consultation clause triggered in June 2023, the Central Bank of Kenya raised
its policy rate by 100bps followed by a further 200bps in December.

A consequence of that mis-prescription and those actions is that once the immediate risk of default had
passed with issue of a new Eurobond at 10.6 percent in February 2024, the exchange rate began
strongly appreciating, once again, even through the recent protests (Figure 24).

Figure 24. Daily Exchange Rate, Kenya Shillings per US$ 2019-2024

And an even more fundamental consequence of the entire episode concerns non-food in ation—
Kenya’s de facto core in ation in the context of the food supply shocks. That, measured over 12
months, remained within the target range throughout the IMF program, averaging close to 5 percent
(Figure 25 blue line). But that number makes no correction the cascade of indirect tax increases
occurring so that the underlying rate relevant for monetary policy is lower—and possibly considerably
so.

37
fl
fl
 Figure 25. Non-Food In ation, 2019-24

And most recently, the rolling three-month average of its one month annualized rate (green line), again
uncorrected for indirect tax increases, has fallen well below the target range for a signi cant period
with almost all of its components trending this way over that period (circled, Figure 25).

Thus, despite the headline in ation rate, monetary policy has erred to over-tight. And after the most
recent set of policy rate rises, that chronic problem has become acute: monthly non-food in ation is
now signi cantly undershooting the target range and absent action looks set to continue doing so.

This last highlights once again the misspeci cation of the IMF consultation clause over 12 rather than
three months seasonally-adjusted annualized, an error guaranteed to engender tardy IMF intervention.

All this directly contradicts the most recent IMF staff statement endorsing the current monetary
stance. Even given uncertainty ahead of the February Eurobond issue, the increases in the Central
Bank Rate from 9.5 to 13 percent—i.e., starting with the rise which took place as the IMF consultation
clause triggered in June 2023—were overdone. The IMF should call for their reversal immediately.

38
fi
fl
fl
fi
fi
fl
 And IMF staff in that statement also mis-report that a decline over recent months in headline 12-
month in ation is partly the result of “lower food prices.“ 12-month food price in ation has indeed
fallen recently, but bar a marginal drop in April more than fully reversed in May, aggregate food prices
have not fallen over any interval. That uncorrected reporting error, on a matter so central to sound
technical analysis of the appropriate monetary policy stance and to Kenyan political economy, appears
to be symptomatic of much deeper aws in IMF staff monetary analysis.

With the major real income consequences of the food price shocks thus unaddressed alongside
excessive monetary tightening, a fundamental strategic error of the IMF in the program was to leave
the originally scheduled timing of the revenue increases unchanged, as if that timing was immutable.

That decision compounded severe unanticipated uncompensated non-permanent-but-enduring income

shocks for households from food prices with permanent adjustments from the slew of tax rises. The
burden was heaviest on those at the low end of the income scale.

The combination has completely broken public support for the scal program—with recent proposed
tax rises on bread, diapers, sanitary towels, motor vehicles, and fuel apparently the straws which broke
the camel’s back.

3. International Food Price Transmission

And the IMF also overlooked key policy issues arising from the composition of the food price shock.

IMF staff treatment of the food price shock as a challenge for in ation—requiring policy focus only on
its second round effects—rather than a what it was, a set of enduring but non-permanent relative price
shocks meant that they failed to identify root causes and hence the appropriate policy response.

Of the four components which dominate the food price aggregate, namely cereals, live animals and
meat, milk and dairy products, and vegetables, the last has seen its prices rise the most, doubling since
2019. That largely re ects drought, which also drove up local meat and milk prices. All that warranted
a temporary targeted income support program for the most affected (poorest) households (Figure 21).

But second to vegetables in price increases—up by 50 percent since 2019—with a signi cantly greater
in weight in the basket are cereals. These, which include nal goods, saw prices mostly track the
Kenya Shilling value of the FAO global index of cereal raw material prices up following the Russian
invasion of Ukraine. But that index of global cereal raw materials has fallen almost 20 percent during
2024, re ecting declines in the US$ index and the appreciation in the Kenya Shilling, without any
corresponding fall in the cereals price index in Kenya through June, contra the IMF staff statement.

The transmission of global raw material cereals prices to the domestic markets of cereals nal goods in
Kenya is, apparently, malfunctioning if it has not actually broken down.

Given the centrality of food prices to the current Kenyan macro, to its political economy, and to the
poorest, it is remarkable that this key matter has apparently not even been raised by IMF staff.

And that is all the more remarkable given their due attention to and insistence that global crude oil
prices are properly and swiftly transmitted to the domestic market for nal fuel prices via the orderly
implementation of the domestic fuel price formula.

39
fl
fl
fl
fl
fi
fi
fl
fi
fl
fi
fi
Even allowing for the distinction between raw material and nal goods and possible systematic
smoothing of noise in raw material prices and the exchange rate, the reason for the recent behavior of
Kenyan cereals consumer prices in the global context is unclear. Obvious suspects include tariff and
non-tariff barriers and other public policy interventions in local cereals markets.

If IMF staff ignored these grateful for the revenue raised, then they have doubly failed on this matter
—prioritizing revenue over the outward orientation and ef ciency of the economy and over the
regressive effects of that revenue, notably given other food price shocks alongside.

Thus, given that the global cereal price US$ index of cereal prices is now broadly unchanged in real
US$ terms from its pre Ukraine levels, rather than these foods compounding the case for temporary
targeted income support in Kenya, they warrant urgent analysis to identify and, as needed, correct
whatever may be impeding the transmission of falling global raw to domestic nal prices now.

4. Warming

Following the election of a new President in September 2022, the IMF program was supplemented in
mid 2023 by IMF credits under its global-warming-related Resilience and Sustainability Facility.

This added a raft of additional conditionality attached to each of its 10 “Reform Priorities”.

But again, key strategic issues underlying the decision to extend this form of IMF nance were simply
not discussed or addressed.

Most basically, it is not explained why these matters were of such immediate macroeconomic priority
as to render access to essential immediate IMF nance conditional upon them.

This consideration does not mean that Kenya should desist from the administrative preparations
related to prospect of further warming. At their sovereign discretion, the authorities are at liberty to
proceed with those. But immediately needed IMF nancing for the macro should not be contingent on
matters which may or may not arise 20 or 40 years from now.

Then, whatever the case for administrative preparations on various matters for Kenya to better
anticipate the hit it will take from the emissions of others, there is no staff discussion of the
implications under the conditionality for Kenya’s own emissions.

Thus, given that Kenya has in the last decade discovered a substantial oil eld in Turkana, and that the
President has indicated in mid-2023 that in light of climate concerns he had decided to “take another
route” rather than develop it, the message to markets is that if Kenya were to discover more or more
plentiful or more remunerative carbon assets, that it would forgo them. That stance runs directly
contrary to immediate need to strengthen investor con dence in Kenya’s macroeconomic program.

Were Kenya to follow through with such commitments to curb its emissions, this would overlook that
the challenge of warming emanates from the global stock of emissions, not the ow, and Kenya’s (and
Africa’s) contribution to the stock is negligible, and would remain negligible even were it and Africa to
exploit their carbon assets to the full. By implicitly forgoing this income, a country in which one in ve
children are stunted would undertake to bear a grossly exaggerated—and likely Quixotic—share of
the burden to address the problem. All these issues simply pass without comment from the IMF.

40
fi
fi
fi
fi
fi
fi
fi
fl
fi
fi
 And of immediate concern are trade agreements, including the free trade agreement with the European
Union, which in these circumstances impose various barriers (including CBAMs) on Kenyan exports
if those exceed the Europeans’ targets for their carbon content. The IMF is simply silent, yet again, on
the de facto rendering of Kenya’s carbon resources as stranded assets, and why on top of the long-
standing fundamental misallocation of land, this represents a warranted additional constraint on
Kenya’s development and on prospects for success of its IMF program.

5. And then, Floods

Yet as the IMF’s principal shareholders nevertheless impose severe emissions constraints on Kenya,
they continue adding to their own hugely disproportionate contribution to the stock of emissions.

One symptom of that was unprecedented rains and ooding nationwide in April 2024—in which well
over 250 Kenyans lost their lives and some 400,000 endured forced evacuations. For many citizens,
this constituted the last straw.

The Kenyan authorities, initially, were caught completely off guard: the president welcomed the rains
as answers to his prayers for droughts’ end and promptly left for three days’ African diplomacy; his top
economic advisor observed that Kenyans die all the time and that government was not their agony
aunt; and recovery has featured discriminatory clearance of people from riparian areas.

If the IMF, after warm words of concern, ends up treating the issue in the latest program review in
similar de facto manner, it risks completely discrediting the program in the eyes of the general public.

The challenge for the IMF is not only to address the further upward jolt to food prices and the hit of
yet-to-be seen scale and duration to output and revenue, but longer term, is also to address the
profound inadequacies in Kenya’s infrastructure, notably drainage, which the oods revealed.

It would be profoundly ironic—to put it mildly—were the IMF to press ahead with its “10
Administrative Reform Priorities” while insisting on above-best-practice primary balance targets
which preclude infrastructure investments which recent ooding has revealed as essential. This would
be a severe case of climate form over substance.

6. Doom Loop

Furthermore, having misspeci ed the original program scal conditionality by setting above best
practice medium-term primary balance goals and unduly backloading the initial scal adjustment, and
having compounded that with an excessively tight monetary policy especially from mid 2023, output
and hence revenues are both systematically falling short of program projections and targets.

Instead of reading those as yet a further signs of misspeci ed conditionality, the IMF has evidently
responded in 2024 by raising both its revenue ratio targets and its demands for further tax increases,
both of which are re ected in the 2024 Finance Bill recently, after major public protest, withdrawn.

Far from xing the problem, this doubling down on the scal side—described, repeatedly by IMF
staff, as “correcting course”—compounds it. It adds further scal withdrawal to that from monetary
policy, depressing demand further. If that is sustained, the revenue base for 2025 will also be further
depressed. And the clearest symptom of this will be a continued signi cant undershoot of annualized
seasonally adjusted one month rates of non-food in ation.

41
fi
fl
fi
fl
fl
fi
fi
fl
fi
fi
fi
fl
fi
 The rst corrective action to exit this tailspin is not to raise revenue ratio targets and associated tax
rates as in the original 2024 Finance Bill—it is a decisive relaxation of monetary policy, without delay.

And the entire preceding discussion brings ever more sharply into question the strategic decision at
program’s outset, never discussed or explained then or since, to shun debt restructuring.

7. Christmas Tree Conditionality

With all of this, the program was seriously overladen with conditionality.

As listed by IMF staff, it started with 26 conditions, breach of any one of which could in principle have
caused disbursements to stop (Figure 26). That number had ballooned by the 6th Review to 36,
including those associated with borrowing under the Resilience and Sustainability Trust which began
with the Fifth Review. As several of these were compound conditions, the effective number of
conditions was typically considerably higher throughout.

Figure 26. Number of Formal Conditions at Origination and at Each Program Review

This does not just represent a failure by the IMF to focus on the macro-critical essentials. It also
represents an undue intrusion into sovereignty and, in principle, it implies that immediately needed

42
fi
 nancing could be stopped in view of considerations arising decades away in some cases. By littering
the program with reasons why it could be stopped, the paramount purpose was compromised—namely
to buttress private investor con dence in Kenya.

And the conditionality does not stop there; it is supplemented by multiple additional requirements
imposed by the dominant IMF shareholders via such as the EU Economic Partnership Agreement
with Kenya and CBAMs.

Partly in regard to such considerations, an informal guideline was adopted by the IMF following the
Asian Crisis programs in the late 1990s—which featured similarly counterproductive excess
conditionality. According to that, no more than 10 conditions should be attached to each program.

That guideline has been comprehensively breached in the IMF program in Kenya, with all the
attendant consequences the guideline was intended to avert. This is the consequence of the IMF going
far beyond its proper role of supporting Kenya with limits on macro policy to implicitly trying to run
it.

And as nal indictment, the blizzard of conditionality clearly failed to spare Kenya its current impasse.

8. Market yield spreads

On their own terms, the classic—indeed foundational—purpose of IMF programs is to catalyze

(sometimes restart) private capital ows. Yield spreads thus provide the litmus test of whether that
objective has been achieved.

With 300 basis points re ecting the differential between the in ation targets of the respective central
banks, long term differentials between Kenya and the US remain well above pre program and even
mid-program through 2023, despite determined efforts by the authorities to comply with all IMF-
critical conditionality (Figure 27).

43
fi
fi
fl
fi
fl
fl
 Figure 27. Yield Differentials, 10 Year Government Bonds, Kenya less US$, 2021-2024

So, far from zealous IMF mandated medium-term scal adjustment being rewarded, it is is re ected in
the high yield differentials and the exorbitant yield required to oat the re nancing Eurobond in
February 2024; 10.6 percent compared to a yield of 7¾ percent for Senegal shortly thereafter.

Accordingly, it is not only the Kenyan citizenry which has delivered its verdict on the IMF program.
Despite the overload of IMF conditionality, given the fundamental misspeci cation of the program and
then its inappropriate—non—adjustment to Ukraine, droughts, and oods, markets are now even
more deeply skeptical than they were at program’s eve.

And that remains so even after re nancing the maturing Eurobond and con rmation that agreement
with staff on the current program review had been secured.

9. Macroeconomic Judgement for 2024/25

To reset the program now, the starting point is that the primary balance outturn for 2023/24 of a de cit
of some 1 percent of GDP was already at the best practice optimum for countries at Kenya’s GDP per

44
fi
fi
fl
fl
fi
fi
fi
fl
fi
 capita. Alongside, there is clear evidence from non-food prices that underlying in ation is now
signi cantly undershooting the Central Bank of Kenya target range.

Accordingly, there is neither a demand-management nor a growth-potential case for further scal
tightening in 2024/25.

Instead, the underlying primary balance for calendar 2024/25 should target no change from 2023/24,
rather than the 1 percentage point of GDP further consolidation originally mandated by the IMF.
Furthermore, there is no demand management or growth potential case to continue the originally
mandated consolidation to 2025/26 or later. It should be cancelled entirely.

In that context, there would be a case to add to the underlying primary de cit of 1 percent of GDP a
temporary targeted program of income support for the poorest to address, nally, the food price shock.

Given all that and continued Shilling misalignment, immediate policy to get underlying in ation back
up to target should be led by a 300 basis point cut in the Central Bank of Kenya policy rate, and by
more when—as seems increasingly imminent—the US Federal Reserve lowers the Federal Funds
Rate.

If, for whatever reason, the Central Bank of Kenya fails to deliver that, there would be a second-best
case for an underlying modest relaxation of scal policy in 2024/25. And if the authorities end up
implementing all or part of the IMF’s originally mandated consolidation for 2024/25, there would be a
second-best case for the Central Bank of Kenya to cut immediately by 400 basis points.

In context of this macroeconomic judgement, the only case for the further scal consolidation that the
IMF originally mandated for calendar 2024/25 and beyond is pursuit of its sub-best-practice medium-
term primary balance target. And the only argument for that would be if best practice medium-term
balances cannot be reconciled with debt sustainability.

In that event, however, given the premium on raising Kenya’s medium-term growth outlook to
potential as indicated by its best performing peers, strategy should not focus on further underlying
scal consolidation from 2023/24 but on pursuit of debt restructuring and write offs anchored by a
medium-term primary balance target of a de cit of 1 percent of GDP. This “macroeconomic audit” of
public debt, including issues of both ex ante and ex post transparency of debt acquisition and steps to
improve expenditure ef ciency, is needed now along with any “accounting” audit to be undertaken.

In light of this macroeconomic judgement and growth and debt strategy, the President was right nally
to veto the 2024 Finance Bill. He subsequently announced that negotiations to replace it and the
associated expenditure estimates for 2024/25 would be based essentially on no change in the overall
scal balance relative to GDP relative to the 2023/24 outturn.

However, all of this is subject to ongoing negotiation, including with the IMF. At this stage, it remains
unclear if the IMF has moved even as far as a merely symbolic partial rephasing of its mandated scal
consolidation for 2024/25 matched by corresponding greater consolidation in later years.

But by diffusing some of the tension, the full agenda proposed above for 2024/25 would provide an
appropriate macroeconomic backdrop against which Kenya can address the fundamental governance
issues arising from the protests and the of cial responses in recent weeks.

45
fi
fi
fi
fi
fi
fi
fi
fi
fi
fi
fl
fl
fi
fi
fi
 IV. Assessment

The IMF was right in early 2021 that Kenya needed a successor program to the Covid Rapid Credit
Facility and that it should be anchored in scal adjustment.

But beyond that, the program was born in mis-diagnosis, notably of Kenya’s underperforming growth,
the exchange rate, the economy’s orientation, and the public debt stock.

It has been bereft of any substantive explanation for its core medium-term primary balance or revenue
parameters, or of the decision to shun public debt restructuring.

Targeted scal adjustment should have been front-loaded to let monetary policy drive exchange rate
correction. But instead it was backloaded and the ultimate target balance was well-above best practice.

Despite the given rationale for rising revenue ratios to support government functions, they have been
primarily in service of those sub-best-practice primary balance targets and have yielded chronic
instability in tax structures.

And the program has been conducted intellectually chaotically by the IMF across its reviews,
including that monetary conditionality was not respeci ed given large supply shocks.

Then mid-stream, a raft of conditionality on warming was bolted onto it without explaining why that
was the macroeconomic priority at that point and not food prices, why Kenya’s carbon assets should be
stranded, or latterly reconciling the entire program with ood-revealed infrastructure imperatives.

And despite obsession with innumerable extraneous details and amid a major enduring food relative
price shock, there has been absolute silence from the IMF on land.

Now, unenumerated by the IMF, real incomes are still falling, plummeting in some cases especially at
the low end, and the program nds itself in a doom-loop—with output and revenue shortfalls
generating tax adjustments which compound output shortfalls, and so on and on to uprising.

After three years and SDR 2 billion of disbursements, Kenya barely scraped oatation of a new
Eurobond to re nance that maturing—but at an exorbitant yield—and broader yield spreads continue
to re ect deep market skepticism. Signs are that the faith of the population has similarly been lost,
notably among the young, heralding deserved generational reputational damage for the IMF in Kenya.

All of this represents a gross failure of IMF internal quality control. And as that has been persistent
and systematic, ultimate responsibility rests with Management and the IMF Executive Board.

But instead of acknowledging and correcting it, the IMF de facto response is exasperation at market
and public disquiet, insisting that Kenya “tough it out” with that indignant tone echoed, perhaps under
duress, by some local of cials.

That response cannot stand. It overlooks that the IMF celebrated the pre-program errors as “Africa
Rising” in the mid 2010s, that vaccine Apartheid and relentless carbon emissions by its main Executive
Board members have greatly aggravated dif culties, that the IMF turned a blind-eye to the opulence

46
fl
fi
fi
fi
fi
fi
fi
fi
fl
fl
 in Kenya that it nanced since 2020, and that while scal adjustment was unavoidable by 2021,
incoherence in IMF-mandated adjustment was and is not.
And “tough it out” just begs the question of how tough is tough enough for them? With one in ve
children stunted, will two in ve be tough enough?

The de ance in that rebuke is no impropriety. While there are times when public protest rails at
painful but necessary macroeconomic adjustment, there are other times when it informs IMF macro-
economists—if they will hear—that they are in error. That is the case now.

In particular, as the original misspeci ed program with its unduly backloaded scal adjustment and
systematically overstated output projections evolved, a further fundamental strategic error of the IMF
was to leave the originally scheduled timing of the revenue increases and trajectory of primary balance
targets essentially unchanged, as if the timing for those was set on tablets of stone.

By thus essentially ignoring the food price shocks, those decisions treated the consequent severe hit to
household income as permanent, combining it with scheduled and doom-loop catch-up permanent
increases in taxes, with the combined toll especially severe on the poorest.

And then, oods.

A critical consequence is that all this has put a reform-minded President into a position where he has
appeared, up until now, to hear the IMF more than he hears Kenyans.

And the upshot of that is that Kenya now has to secure a belated fundamental economic reorientation
as well as to resolve complex debt sustainability and governance issues simultaneously.

That is a challenging trio because the intended boost to business investment from reorientation will be
discouraged by the dif culties inherent in resolution of the debt and the political challenges.

The answer is not to toss the IMF baby out with the bathwater—but it has to change tack, abruptly.

The rst steps to reset the program are an immediate decisive relaxation of monetary policy, to
respecify the IMF consultation clause onto non-food in ation, and a downward revision of the
medium-term primary surplus target to best practice.

In that context, the apparent withdrawal of the mandated scal adjustment for 2024/25 of 1 percentage
point of GDP is welcome. But that is just a half step until the IMF approves it and withdraws the
requirement for subsequent mandated scal adjustment. That full package would set the stage for an
immediate resumption of outward-oriented growth with in ation on target and combat protectionist
pressures, including on foods.

On top of that, there is need and scope for a temporary targeted program of income support for low
income households in respect of the still ongoing food price shock, accompanied by action to accelerate
the transmission of global cereals prices to the domestic market.

If, as possible, all of that is insuf cient to quell market disquiet or to reconcile debt sustainability with
best practice medium term primary balance targets or to calm public protest, an immediate review by
the IMF of the necessity and quanta of debt write-offs should be conducted—fully re ecting the

47
fi
fi
fl
fi
fi
fi
fi
fi
fi
fi
fl
fi
fl
fi
fl
fi
 impact of corrected misalignment on the debt stock, and anchored in lowering the medium-term
primary balance target fully to best practice, namely to a medium term de cit of 1 percent of GDP.

And the number of IMF conditions has to be cut drastically. Apart from the loss of focus, intrusion
into sovereignty, and compromised message to markets from current requirements, the IMF team
managing this program give every impression of being overwhelmed by the sprawling number of tasks
they have assigned themselves—running across hundreds of State Owned Enterprises amongst many
other matters. So when the macroeconomic ground shifts beneath them mid-course, they simply do not
have the band-width to discern the implications and re-direct the program core as needed.

To x that, the IMF should discontinue lending to Kenya from the Resilience and Sustainability
Facility leaving the authorities to address those issues at their sovereign discretion, should replace the
associated nancing from the Extended Fund or Credit Facilities correspondingly, should specify
aggregate scal conditionality only on the primary balance, and it should at global level loudly
challenge CBAMs and such imposed by its dominant shareholders precisely because they penalize the
countries, like Kenya, which even collectively are not responsible for the global stock of emissions.

And for the remainder of this program and any successors, the IMF should stop concealing the
rationales for core strategic decisions and parameters—including the medium-term primary balance
and immediate revenue ratio target numbers. Doing so substantively breaches IMF policy on
transparency, under which only matters potentially disturbing to markets may be withheld (and only
at the authorities speci c request). It is also poor professional practice given that full disclosure is the
last best assurance against serious technical and strategic error. And it is an unfortunate echo of times
past when the fates of African populations were similarly determined elsewhere, without explanation.

More broadly, the IMF has often implicitly excused its second-best conditionality behind the mask
that programs were “home grown”—with the unstated implication for such as Kenya that the
authorities made the choice against debt write-offs.

But such authorities design their “home grown” agendas in a context—in this case the dysfunctional
arrangements to resolve Sovereign Insolvencies approved collectively by the dominant members of the
IMF Executive Board. Those constraints imposed on home-grown options from 2021 onwards
represents the most fundamental failing of IMF operations in Kenya since then.

And the consequent social unrest now represents yet a further and increasingly severe indictment of
those arrangements. That unrest should further motivate immediate fundamental reforms to the global
sovereign insolvency architecture to anchor it in best practice medium-term primary balances.

There were no signs in the latest IMF staff statement that any of these adjustments were forthcoming
until, from June, Kenyan youth took their exasperation from their phones onto the streets. Following
that, the IMF like others has to consider its position. It has given no public sign to this point that it will
even go as far as merely rephasing some of its mandated sub-best-practice primary balance targets for
2024/25 to subsequent years, and requiring that remaining losses of revenue from planned 2024/25 tax
increases be fully matched by cuts to planned budgets.

If that is all it does, while leveraging protest to curb low quality spending is welcome, that leverage
should not be used to deliver macroeconomically mis-speci ed budget balance targets beyond 2024/25.
If it is, not only will policy remain sub-best-practice but the social turmoil will simply ratchet up again
next year.

48
fi
fi
fi
fi
fi
fi
Accordingly, President Ruto should not regard advocacy of the full agenda proposed here either as
ingratitude for what the IMF has done so far nor as extending yet another supplicatory black hand,
but rather as Kenya—and Africa—demanding to be treated properly.

Nor should he, on Kenya’s behalf, be resigned to half gestures from the IMF nor see them as some
morality tale consequent the policy missteps of earlier administrations. Ultimately, there is no virtue or
African pride in paying debt to the point of hunger and disorder, and absolutely no excuse—with
dozens of young protesters dead and many more disappeared—for the IMF to fuel those ames.

Their doing so has become a major impediment to Kenya taking its rightful place as one family, one
nation, one language—on long-run growth as in all long distance races—out in front.

49
fl