Mexico Needs A Fiscal Twist: Response to Covid-19 and Beyond
Summary
IMF Working Paper WP/20/215, “Mexico Needs a Fiscal Twist: Response to Covid-19 and Beyond,” by Swarnali Ahmed Hannan, Keiko Honjo and Mehdi Raissi, dated October 2020. The abstract says Mexico’s fiscal response has been modest compared with peers, reflecting the authorities’ wish not to issue new debt for spending, and risks a more severe recession. The paper proposes announcing a pro-growth, inclusive medium-term fiscal reform upfront to open space for larger short-term support, and the simulations use the IMF’s GIMF model. The introduction puts additional expenditures at 0.2 percent of GDP for health care and 0.5 percent of GDP to protect households and firms.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Archived web capture. Preserved source — captured from www.imf.org; archived 2026-09-23.
Cite this document
“Mexico Needs A Fiscal Twist: Response to Covid-19 and Beyond.”
Original at www.imf.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/wpiea2020215-print-pdf-ashx-1694d696731cfd17-80fc3255e9/
, SHA-256 420e75046910ab256c265aa091ab3f868cdef683266266c49b74d57e6bc5eee9.Full text
WP/20/215
Mexico Needs A Fiscal Twist: Response to Covid-19 and
Beyond
by Swarnali Ahmed Hannan, Keiko Honjo, Mehdi Raissi
IMF Working Papers describe research in progress by the author(s) and are published
to elicit comments and to encourage debate. The views expressed in IMF Working Papers
are those of the author(s) and do not necessarily represent the views of the IMF, its
Executive Board, or IMF management.
© 2020 International Monetary Fund WP/20/215
IMF Working Paper
Western Hemisphere Department
Mexico Needs a Fiscal Twist: Response to Covid-19 and Beyond
Prepared by Swarnali Ahmed Hannan, Keiko Honjo, Mehdi Raissi
Authorized for distribution by Rishi Goyal
October 2020
IMF Working Papers describe research in progress by the author(s) and are published to
elicit comments and to encourage debate. The views expressed in IMF Working Papers are
those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board,
or IMF management.
Abstract
Mexico’s fiscal response to the pandemic has been modest compared to its peers, reflecting
the authorities’ desire to not issue new debt for spending. This approach, however, risks a
more severe recession and a weaker economic recovery, with further costs in the future.
Balancing the need for stronger near-term fiscal support for the people and the recovery
against medium-term discipline, this paper lays out an alternative strategy. We show that
credibly announcing a pro-growth and inclusive medium-term fiscal reform upfront—
including increased tax capacity, higher public investment and strengthened social safety
nets—would open space for larger short-term support and close medium-term fiscal gaps.
Model simulations suggest that this package would boost output, limit lasting economic
damage from the pandemic, and put debt trajectory on a declining path in the medium term
as tax reforms pay off and risk premia decline.
JEL Classification Numbers: H2, H5, H6, E2, E6.
Keywords: Covid-19, fiscal response, tax, social safety nets, general equilibrium model.
Author’s E-Mail Address: sahmed@imf.org, khonjo@imf.org, mraissi@imf.org
3
I. INTRODUCTION1
Mexico’s fiscal response to the Covid-19 shock has been modest. Thus far, additional
expenditures have amounted to 0.2 percent of GDP to support health care and 0.5 percent of
GDP to protect households and firms (see Box 1 for details). Loans to formal workers and
recently laid-off employees as well as contingent liabilities added up to 0.5 percent of GDP.
As illustrated in Figure 1, the level of support is significantly lower than the peers in Latin
America and the G20, and may lead to economic scarring. The size of spending and revenue
measures in 2020 is also smaller than the announced discretionary fiscal stimulus in 2009
(Figure 2), which included public investment and targeted transfers.
Figure 1. Fiscal Measures in Response to Figure 2. Emergency Lifelines in 2020
the Covid-19 Pandemic (Percent of GDP) versus Announced Discretionary
Stimulus in 2009 (Percent of GDP)
Sources: National authorities and IMF staff estimates. Source: FM Measures Database and SDN/15/06.
Note: Weighted by GDP in PPP-adjusted current US dollars for
country groups. Revenue and spending measures exclude deferred
taxes and advance payments. See Fiscal Monitor Database of Covid-
19 Measures for details.
Unlike a typical macroeconomic disturbance, the Covid-19 pandemic and policies
implemented to contain its spread have brought about simultaneous disruptions to demand
and supply in a totally new economic environment. Chudik and others (2020) quantify the
economic impact of the pandemic on various countries/regions and show that these effects
could be long-lasting without decisive policy actions. They highlight the importance of
policy interventions that can restore the normal functioning of financial markets, as well as
adopting fiscal measures that can limit bankruptcies of viable firms and support incomes of
households. Similarly, IMF (2020a) shows that the announced fiscal policy measures in Latin
America and the Caribbean are playing an important role in mitigating the effects of the
pandemic. Moreover, Deb and others (2020) find that while containment measures have led
to a large impact on economic activity, fiscal and monetary policy measures used in response
to the crisis were effective in mitigating some of these economic costs.
1
The authors would like to thank Rishi Goyal for his overall guidance; Juan Pablo Cuesta Aguirre for excellent
research assistance; Javier Ochoa for excellent editorial assistance; Krishna Srinivasan, Alejandro Werner,
Nikolay Gueorguiev, Michael Perks, Francisco Roldan, Saji Thomas, and colleagues from IMF, Secretaría de
Hacienda y Crédito Público (SHCP, Mexico) and Banco de México for their helpful suggestions and comments.
4
To evaluate the case for fiscal action, it is helpful to consider three phases of the Covid-19
pandemic (Figure 3). The first is the acute phase in which lockdowns are fully in place. The
role of fiscal policy in this stage is to save lives and livelihoods by relying on three levers
(see IMF 2020b for details): (1) fully accommodating spending on health care; (2) providing
emergency lifelines2 to households and firms to cushion the adverse impact of lockdowns,
including in hard-to-reach informal sectors; and (3) letting automatic stabilizers operate,
especially unemployment benefits and social safety nets. The second phase is when
lockdowns are gradually lifted under uncertainty about the course of the Covid-19 pandemic
and the economy. In this phase, the role of fiscal policy is to gradually unwind emergency
lifelines while strengthening social safety nets in a more durable way, supporting the
resumption of economic activity, lifting demand with fiscal stimulus, and, if needed,
facilitating a reallocation of resources (rather than saving unviable firms). The final phase is
when pharmaceutical interventions and medical treatments are widely available, and the
pandemic is under control. In this period, fiscal space should be re-built through growth-
friendly and inclusive adjustments.
Figure 3. The Role of Fiscal Policy at Different Stages of the Crisis
Pandemic under
Great Lockdown Gradual Reopening
Control
Number of new cases
Healthcare Response
Containment and mitigation measures
Emergency Lifelines
Support for affected people and firms
Existing Automatic Stabilizers:
Progressive taxes, unemployment benefits, SSNs
Fiscal Stimulus
Fiscal Structural
Reforms
Time
In this paper, we use the above three-phase schematic to elaborate the elements of a
comprehensive fiscal policy response to Covid-19 for Mexico. We also simulate, using the
IMF’s Global Integrated Monetary and Fiscal (GIMF) model—see Kumhof and others
(2010) and Anderson and others (2013) for modeling details—the macroeconomic effects of
a proposed package that comprises larger fiscal support in the short-term and a growth-
2
The difference between lifelines and fiscal stimulus is that lifelines are aimed at sustaining basic consumption
levels by supporting households and helping viable firms temporarily (e.g., through expanded unemployment
insurance, enhanced social safety nets, and liquidity support). Fiscal stimulus aims at boosting aggregate
demand through broad-based measures (e.g., public consumption and investment, tax cuts).
5
friendly and inclusive adjustment over the medium term. This is akin to a “fiscal twist”.3 The
model simulations suggest that a medium-term tax reform, alongside permanently higher
capital spending and enhanced social safety nets, can generate sizable output gains, more
redistribution, and a sustainably lower public debt ratio over the long term. Short-term output
costs of the tax reform, if implemented credibly, are expected to be limited.
The rest of the paper is organized as follows. Section II discusses the desired near-term
policy response to Covid-19 (for phases 1 and 2) and their cost estimates. Section III presents
options for phase 3, including a medium-term tax reform. To enhance the political and public
acceptability of the reform package, it also explores ways to enhance social safet nets.
Section IV quantitatively simulates the macroeconomic effects of a growth-friendly mix of
spending and revenue measures. Section V concludes.
Box 1. Government’s Fiscal Response to Covid-191
In response to the Covid-19 pandemic and its economic fallout, the Mexican government has
implemented a range of measures: 1) frontloading two bimesters payments of the old-
age and disability pensions; 2) accelerating procurement processes and VAT refunds; 3)
lending to firms and workers in both formal and informal sectors; 4) providing liquidity
support through development banks.
Overall, the above-the-budget-line fiscal measures amount to 0.2 percent of GDP in health
spending and 0.5 percent of GDP to support households and firms. Below-the-budget-line
measures amount to 0.5 percent of GDP comprising loans to formal workers and recently
laid-off employees as well as contingent liabilities.
Health spending. The government’s measures aim to ensure that the Ministry of Health has
sufficient financial resources and does not face red tape in procuring medical materials.
Additional spending.
• The Ministry of Economy is granting loans with optional repayments to 1 million small-
and medium-sized enterprises (SMEs) that maintain employees on payroll, self-
employed and domestic workers (25 thousand pesos per loan), and another 1 million
loans to family businesses previously registered in the Welfare Census
(25 thousand pesos per loan).
• The government is providing subsidized unemployment insurance for 3 month to
workers that hold a mortgage with the Housing Institute (7.3 billion pesos).
3
Note that the global nature of the Covid-19 pandemic and the synchronicity of its macroeconomic effects
might justify tolerance for higher debt levels temporarily. Chudik and others (2017) argue that there is no
compelling evidence for a universally applicable threshold effect in the relationship between public debt and
growth if one accounts for global interdependencies and common factors/shocks. However, they argue that a
persistent accumulation of government debt is associated with lower growth in the long term. This suggests the
need for medium-term fiscal adjustments once the pandemic is under control and the recovery is firmly in train.
6
• Additional resources are allocated to social spending related to infrastructure, security,
education, and other areas (50 billion pesos).
Accelerated spending.
• The government has frontloaded social pension payments for the elderly and disabled
people by 4 months.
• Procurement processes and VAT refunds are to be accelerated.
Below-the-budget-line measures and contingent liabilities.
• The Government is implementing other measures, including housing credits for
government workers with low-interest rates (ISSSTE’ loans for a total amount of 35
billion pesos) and personal loans at a low rate (3 billion pesos).
• Development banks are to extend loans amounting 64.5 billion pesos, largely to SMEs.
Mexico: Key Fiscal Measures in Response to COVID-19
Amount
Main measures (percent of 2020
GDP)
Total 1.1
Measures with budget impact 0.7
Support for households 0.0
of which
Unemployment subsidy for 3 months to workers that hold a mortgage with the Housing Institute (MXN 7.3bn) 0.0
Support for businesses 0.2
of which
Loans with optional repayment to SMEs that maintain employees on payroll, self-employed, and domestic workers 0.1
Loans with optional repayment to family businesses previously registered in the Welfare Census 0.1
Supporting the health system 0.2
Others (additional social spending) 0.2
Below-the-line measures 0.5
Institute for Social Security and Services (ISSSTE) loans to state workers 0.2
Personal loans by the Institute of the National Fund for the Consumption of Workers (Fonacot) (MXN 3bn) 0.0
Loans by development banks, particularly to SMEs 0.3
1
Sources: Mexican authorities, staff estimates, IMF policy tracker, and IMF Fiscal Monitor: Database of
Country Fiscal Policy Measures in Response to the Covid-19 Pandemic.
II. DESIRED (NEAR-TERM) POLICY RESPONSE TO COVID-19
Whereas in normal economic downturns a key goal of fiscal policy is to stimulate demand,
this crisis is like no other. In its early stages, the primary policy objectives are to: (1) save
lives by containing the spread of the disease and treating those who are infected; and
(2) protect livelihoods from the economic fallout of Covid-19. Correspondingly, the first
policy priority is to fully accommodate spending on health care. The second priority is to
adopt timely, temporary, and targeted fiscal actions to protect people and firms, including in
hard-to-reach informal sectors. Such support is likely to provide the most effective cushion to
7
output and essential consumption because it alleviates the drop in incomes for people with
limited savings and reduces the likelihood of bankruptcies. In so doing, these actions would
prevent a health crisis from generating large economic scarring effects.
The extent of fiscal actions and the choice of instruments during phase 1 of the crisis should
be guided by some overarching principles, discussed in IMF (2020b). Lifelines should be
cost-effective and transparently embedded in medium-term budget frameworks. Where
possible, measures should build on existing programs to facilitate timely support to
households and firms. Finally, financing constraints will determine the extent of emergency
lifelines in phase 2 and a broad-based fiscal stimulus.
Guided by these overarching principles, it is advisable that Mexico’s near-term fiscal support
to combat the Covid-19 include:
Healthcare
• Universal free coverage for Covid-19 treatment, with no out-of-pocket expenditure, so
that cost or access are not insurmountable barriers to combating Covid-19.
Support to households
• Assistance to households using existing social safety nets and administrative
infrastructure (option 1): support to households could be provided by drawing on a
recent social census covering about 20 million households (out of a total of 36 million)
and Sistema de Información Social Integral (SISI), which collects data from social
programs at all levels of government. Given the size of the shock, the generosity of
existing programs and their coverage could be increased by including informal workers
or those at risk of poverty, and temporarily easing eligibility requirements. The national
voter ID database could be used to identify informal workers who do not have access to
existing social safety net programs—i.e., adults not in the social census nor in SISI nor
in the social security system nor in the tax administration records. Modes of delivery of
assistance could include existing social safety nets, lower utility bills, mailing checks,
setting up bank accounts for direct deposits (which also promotes financial deepening),
or fintech instruments (e.g., mobile/online payments).4
• Assistance to households using enhanced social safety nets (option 2): an alternative but
administratively more challenging approach is to create a single registry of beneficiaries
by merging SISI with the social census. This would enable the government to better
identify those in need, increase cost efficiency, and reduce errors of inclusion and
exclusion, beneficiary overlaps, and program duplications (IMF 2019a).
• Other income and liquidity support measures for formal workers: These include
(i) deferring employees’ social security contributions; (ii) deferring tax payments;
(iii) deferring and/or reducing mortgage, consumer loan, and utility bill payments;
4
Complementary targeted subsidies for health, transportation, utilities etc. for certain sectors (e.g., trade,
manufacturing, construction) or hard-hit regions are also possible options, although not essential if, as in
Mexico, the social safety net has relatively wide coverage.
8
(iv) providing government-funded sick and family leave; (v) relaxing requirements for
accessing social security savings for formal workers—allowing them to draw a fraction
of their future savings from their pension fund accounts on evidence of current hardship.
Similarly, consideration could be given to allowing temporary access to insurance
products, such as annuities. Careful calibration is essential to ensure any future costs are
minimized, including to safeguard the intertemporal sustainability of pension and
insurance funds.
Support to firms
Supporting viable5 firms, particularly small- and medium-sized enterprises (SMEs), could
take two general forms: (i) liquidity support; and (ii) cost-reduction assistance.
• Options for liquidity support include tax and social security contribution deferrals;
accelerated payments—expedited processing of VAT refunds; and clearing public sector
arrears to all suppliers (if needed, prioritize those that use relatively low-skilled workers
intensively)—and loans and/or credit guarantees. Regarding the latter, the government
can (i) provide guarantees for rolling over of bank loans to ensure credit lines are
maintained and essential liquidity remains available. Credit risk should be shared
between the government and commercial banks to align incentives for loan provision
only to viable firms; (ii) lend directly against collateral; (iii) lend to SMEs, with cross-
checks of profitability based on previous tax payments; and (iv) ensure a mechanism is
in place to monitor and record fiscal risks—e.g., to avoid abuse, a central body of
guarantee organizations could check all loan agreements for compliance and guard
against duplication of requests.
• Options for cost reduction include (i) lower social security contributions; (iii) forgivable
loans; and (iii) wage subsidies. One possibility is to provide wage subsidies on
conditions of retaining employees on the payroll to the extent possible while allowing
for lower compensation, and of not using the subsidies for dividends, stock or debt
buybacks, or CEO bonuses. Another option to support firms is to extend the provision of
wage subsidies to informal workers (if any) as well should they be converted to formal
workers. Pre-pandemic, about 12 million informal workers were employed in formal
firms, or close to ⅕ of the workforce. Since much of the benefits would take the shape of
future income protection, that itself would sharply reduce present uncertainty for many
families. Since the trade-off between current support and permanent costs into the future
may be too much for some firms—an issue that needs to be investigated empirically to
discern expected take-up—as an alternative, firms that formalize work arrangements
could receive progressively greater levels of support. That is, the magnitude of
additional support could be conditional on the fraction of informal workers that are
formalized. Note that the provision of wages subsidies to formal workers should be
exempt from this particular type of conditionally.
5
Viability of a firm compares the present value (PV) of profits to the costs of recreating the firm from scratch.
Solvency compares the PV of profits to the value of debt. Debt overhangs from crises can make a firm
temporarily insolvent but still viable.
9
However, for both the liquidity and cost-reduction options, some of the potential negative
externalities should be borne in mind. Deferring social security contributions (SSC) could
defray some burden on formal sector workers in the short-term, but this could also affect the
viability of an already underfunded pension system if continued for an extended period.
Given the high number of infections in Mexico, the depth of the recession to date, and the
relatively weak recovery anticipated by many analysts, the government could consider
keeping emergency lifelines in place for longer and durably strengthening social safety nets
over time. To facilitate the recovery, the authorities could plan to enact, for example,
temporary cost-reduction incentives to encourage private investment, as well as to accelerate
high-return public investment projects, repair, and maintenance activities.6 As uncertainty
about medium-term prospects is resolved, they could also facilitate structural
transformation—facilitating the reallocation of resources toward activities that are likely to
see growing demand once the pandemic is under control—through hiring subsidies
(conditional on formalization) and spending on worker re-training.
A. How much does the near-term policy response cost?
The near-term fiscal response to Covid-19 are grouped into the following categories and
costed: (i) additional health expenditure amounting to 0.6-1.5 percent of GDP; (ii) support to
households of about 1.2 percent of GDP; (iii) support to firms (including wage subsidies of
0.4 percent of GDP); and (iv) increased investment of 0.4 percent of GDP. Overall, the
estimated cost of these emergency lifelines would be 2.5-3.5 percent of the 2020 GDP. This
number excludes support in the form of loans, equity injections, and credit guarantees.
6
General tax rate cuts and holidays should be avoided.
10
Table 1. Estimated Cost of Key Measures
Estimated cost
Measures Assumptions
(percent of GDP)
Total 2.6-3.5
Healthcare 0.6-1.5
Based on Dudine et. al. (2020) that combines a simple
Susceptible-Infectious-Recovered (SIR) epidemiological model
Accommodate additional health spending 0.6-1.5
and assumptions about pre-pandemic spare capacity, cost of
providing care and of increasing capacity in the health sector
Support to households by… 1.2
…expanding existing social safety nets
Offer benefits to those who were eligible but did not receive any
To readily eligible beneficiaries 0.6
social assistance in 2019
3-month support of 2036 pesos per month (average of poverty
To people at risk of poverty lines) for people at risk of extreme poverty, as estimated by 0.3
CONEVAL
3-month support of 2036 pesos per month to informal workers
To informal workers in hard-hit sectors in construction, trade, and services sectors using sector-specific 0.3
lockdown assumptions
Support to firms 0.4
6-month support of 2036 pesos per month to IMSS-insured
Wage subsidies workers of formal firms in construction, trade, and services 0.4
sectors using sector-specific lockdown assumptions
Increase good-quality public investment to support a fraction of
Boost public investment 0.4
private investment lost because of COVID-19
Cost estimates of healthcare Figure 4. Additional Health Spending
in Response to Covid-19 (Share of GDP)
Health spending should be increased to 2.0 (Share of GDP)
mitigate the spread of Covid-19 and treat 1.6
those that are infected (Dudine and 1.2
others 2020). The estimated cost would 0.8
be contingent upon a host of factors: the 0.4
0.0
pre-Covid-19 capacity in the health Saudi Arabia
United Kingdom
United States
Canada
Brazil
system, the cost (both fixed and variable) Japan
Colombia
Chile
Germany
Romania
Australia
Indonesia
Peru
required to expand the existing capacity South Africa
Italy
Spain
Poland
France
Philippines
Turkey
Korea
if needed, and the pathway of the Covid- Russia
Mexico
India
China
19 pandemic. If new cases and hospital Source: Compiled by IMF Fiscal Monitor (2020).
occupancy rates decline durably, health
resources would need to focus on the next phase of the pandemic, such as preparing for a
large-scale vaccine rollout. The cost of prevention and treatment would depend on country-
specific factors such as demography, geography, preparedness of the existing health
infrastructure, effectiveness of non-pharmaceutical mitigation measures, specific factor costs
(e.g., compensation of health professionals), and the stage of the outbreak (IMF 2020c).
11
The estimation of health spending is an Figure 5. Public Health Expenditure
inherently difficult task in the current (Share of GDP)
environment and is fraught with 14
12
uncertainty—related to the trajectory of 10
cases and measurement accuracy, the 8
6
availability of granular data on health 4
capacity, the impact of containment 2
0
measures, economic reopening Germany
United Kingdom
Japan
Canada
Italy
Australia
plans/status, and vaccine development and France
Spain
Korea
Colombia
Hungary
Chile
Bolivia
Poland
Turkey
rollouts. It is even more difficult to Brazil
Russia
China
Vietnam
Mexico
Malaysia
incorporate the quality of healthcare when Philippines
Indonesia
India
Most recent year available. Source: IMF (2019) for Mexico; WDI (via IMF
estimating the cost; this paper does not EAT Database) for all other countries.
explicitly account for this factor. With
these caveats in mind, we consider that health spending could be increased by 0.6-1.5 percent
of GDP in Mexico, based on the following:
(1) Benchmarking: The average health spending on Covid-19 in G20 countries and Mexico’s
regional peers is 0.6 percent of GDP, higher than Mexico’s current Covid-related health
spending of 0.2 percent. Mexico is likely to require even more resources than its peers, given
that its healthcare system is fragmented, service delivery is unequal (spending remains
concentrated on richest states), and administration costs are high (see IMF 2019 for details).
Moreover, health expenditure as a share of GDP and pre-existing number of beds and
physicians are lower than many other countries (Figures 5 and 6). In addition, co-morbidity
factors are high because of elevated incidence of obesity and diabetes, among others.
Figure 6. Pre-Covid-19 Health Capacity
Hospital beds (per 1,000 people) Physicians (per 1,000 people)
16 4.5
14 4.0
3.5
12
3.0
10
2.5
8 2.0
6 1.5
4 1.0
2 0.5
0 0.0
Japan
Korea, Rep. Germany
Italy
Germany
Russia Spain
Russia
Hungary
France Argentina
Australia
Poland
Argentina France
Hungary
China
Australia United Kingdom
Canada
Italy
Spain Japan
Poland
United Kingdom
Canada Korea, Rep.
Mexico
Turkey
Vietnam Brazil
Colombia
Brazil
Chile China
Turkey
Malaysia
Colombia Bolivia
Malaysia
Mexico
Indonesia Philippines
Chile
Bolivia
Philippines Vietnam
India
India Indonesia
Most recent year available. Source: WDI.
(2) Model estimates: We follow Dudine and others (2020) in estimating the size of additional
health spending needed to treat Covid-19 patients based on a simple Susceptible-Infectious-
Recovered (SIR) epidemiological model that projects country-specific number of persons
requiring hospitalization. This number is combined with assumptions about capacity
constraints in the health sector, information about average healthcare cost , and additional
12
funds needed to expand health capacity to compute the overall additional health spending
needed for Mexico. The model captures some non-linear effects: (i) the pressure on the
health system depends more on the distribution of Covid-19 patients than the total number of
cases; (ii) the additional spending needs depend on capacity constraints (spare beds,
ventilators, and availability of doctors/nurses) and technical constraints to health system
expansion; and, (iii) the role of other policies in slowing the spread of the virus.
The estimated healthcare costs are based on two scenarios, detailed in Table 2. These
parameter values can be plugged into the template of Dudine and others (2020) to arrive at
the healthcare cost estimates for Mexico. While the second scenario uses the paper’s default
parameters, the first scenario modifies a few of these values, including reproduction rates (ρ0
and ρ1)—informed by the studies of IMF (2020d) and Parr (2020). The reproduction rates for
the first scenario is comparable to the upper bound (97.5th percentile) of IMF (2020d)’s
estimates as officially reported fatalities could be underestimated by around three times
(Zavala and Despeghel 2020). The first scenario assumptions result in new cases of 16
percent of population by the first year, while the equivalent new cases in the second scenario
is 53 percent of population. When estimating the healthcare costs, the first scenario assumes
a capacity increase of 25 percent maximum (parameter cap in Table 2).
Based on these assumptions, the estimated cost could range between 0.7-1.7 percent of GDP
for total health cost and 0.6-1.5 percent of GDP for public health cost (assuming different
public versus private split, but more than 85 percent in both cases presented). The wide range
of estimates reflects the considerable uncertainties associated with this exercise. In particular,
the cost is very sensitive to assumptions such as how much capacity can be increased and the
reproduction number for Covid-19 (average number of new infections per infected)—factors
that cannot be accurately pinned down ex ante.
Table 2. Epidemiological Parameters
Parameter Scenario 1 Scenario 2 Comment
Share of infected people that is not infectious after 1 week, either because of recovery or
K 0.4 0.4
death.
Average number of new infections per infected person over the duration of infection, when
ρ0 2.4 2.3
no control measures are taken
Average number of new infections per infected person over the duration of infection, when
ρ1 1.3 1.5
control measures (quarantine, lockdowns, etc.) are taken
W 8 10 Number of weeks it will take to switch to ρ1 since the beginning of the outbreak
S 0 0 Share of population that will not get infected (in other words, immune population)
α 1 1 Share of those hospitalized who cannot infect others
Scale up z 1.1 1.1 Multiply estimate of country specific z
Scale up H
mortality
1.1 1.1 Multiply estimate of country specific δ
RNH
Δδ 0.03 0.03 Increase in death rate for people requiring but not obtaining hospitalization
av 0.05 0.05 Share of existing beds at time zero, available to Covid-19 patients
Share of patients requiring but not obtaining hospitalization that can be hospitalized next
nb 0.1 0.1
period
cap 1.25 1.5 Multiple of beds available at time zero above which the health system cannot go
cm US$5,000 US$5,000 Cost of medicine and other material per week, per hospitalized patient
cnb US$25,000 US$25,000 Cost of a new bed and equipment
Fc_100 US$1,000,000 US$1,000,000 Fixed costs associated to expanding capacity by 100 new beds
Source: Dudine et al. (2020) and authors' estimates.
13
Cost estimates of support to households
Figure 7. Social Assistance Spending
(Share of GDP)
Mexico currently spends about 2.2 percent of
2.5 (Share of2019
GDP) 2020
GDP on social safety nets (or social assistance). 2018
2
There are around 8,000 social protection
1.5
programs at the federal, state, and municipal
levels.7 These programs cover 22.9 million 1
Mexicans in the year 2020 (equivalent to 16 0.5
million households or 44.5 percent of total). 0
Those in need are largely identified from the
social census and SISI. The one-year cost of
increased households assistance is estimated at
Source: SHCP
1.2 percent of GDP.
Expanding the existing social safety nets
Estimate 1: Increase the coverage of social safety nets to readily eligible individuals. These
are individuals who have been deemed eligible but have not received any benefits yet. Using
different assumptions about the extent of coverage of such beneficiaries, CONEVAL, which
is Mexico’s agency for measuring poverty and evaluating social development policies,
estimated in May 2020, the additional cost of expanding existing programs to be 0.2-0.6
percent of GDP, based on two scenarios:
• The first scenario assumes full coverage of 8,873,924 beneficiaries who were eligible but
did not receive any assistance in 2019. The additional cost would amount to MXN 137
billion or 0.6 percent of GDP.
• The second scenario assumes a 30 percent increase in coverage (2,815,133
beneficiaries)—the increase in the percentage of the population in poverty between 2008
and 2010—yielding an estimated cost of around 0.2 percent of GDP.
Estimate 2: Increasing the coverage of social safety nets by including those at risk of poverty.
Based on various assumptions about the extent of income loss from Covid-19, CONEVAL
estimates that the total number of people in extreme poverty might increase between 6.1 and
10.7 million, while the UN/ECLAC projects a substantial increase of 14–21 million
(assuming the population living in extreme poverty increases from 11.1 percent in 2019 to
14.9-17.1 percent in 2020). The Centre for Educational and Social Studies estimates that 11.5
million people could move out of the middle class into relative poverty. The economic shock
accompanying the pandemic therefore significantly expands those in relative and extreme
poverty.
7
The IMF defines social protection spending to comprise social insurance and social assistance programs.
Social insurance aims at protecting households from shocks that can adversely impact their incomes and welfare
and is typically financed by contributions or payroll taxes. This includes the pension system and health care for
the population in general. Social assistance aims at protecting households in poverty and is financed by general
government revenue. This relates specifically to poverty alleviation and includes direct and indirect transfers to
poor households as well as social pensions to the elderly to tackle old age poverty. The terms social assistance
and social safety net are used interchangeably in this paper.
14
Using the CONEVAL estimate, the cost of providing temporary support to 6.1-10.7 million
people is 0.2-0.3 percent of GDP assuming that: each individual is provided an income
support of MXN 2036.12 per month (average of urban/rural poverty and extreme poverty
lines, as computed by CONEVAL for July 2020) for 3 months. The estimated cost would be
contingent on the amount of monthly support, its duration, and the underlying assumption on
the duration of the shock. With potentially substantially higher durations of non-employment
and under-employment, the amount of support would be notably higher. For example, using
the same underlying assumptions for computation, 6 months of support would amount to 0.3-
0.6 percent of GDP, while 12 months of support would amount to 0.7-1.2 percent of GDP.
Estimate 3: Increase the coverage of social safety nets to informal workers. Informality in
Mexico’s labor market takes many forms, including those at non-agricultural informal firms,
self-employed agricultural firms, unpaid workers, non-salaried workers8, and workers
without access to social security health services (Alvarez and Ruane 2019). None of the
workers under this definition has access to Mexican Social Security Institute (IMSS). It must
be noted that informal workers could be employed in both formal and informal firms9. There
are about 56 million workers, of which about 23.5 million are formal using the shares from
Alvarez and Ruane 2019 (around 21 million insured by the Mexican Social Security Institute,
IMSS). Given widespread informality, it is necessary to have different programs to target
different forms of informal workers and formal employees.
Covering informal workers of hard-hit sectors depends crucially on the phase of the crisis. In
phase 1, manufacturing, construction, trade, and services were particularly hard hit. In phase
2, auto manufacturing and construction sectors were deemed essential, and a recovery in
production and employment have been underway. However, the loss in employment in
construction was particularly sharp, and is well below pre-pandemic levels. Keeping these
considerations into account, the analysis assumes support for informal workers of
construction, trade, and services sectors. It must be noted that we use high-level assumptions
to show illustrative calculations based on possible strains in certain sectors. These need to be
complemented with analysis using more granular data for policy implementation.
Calculation of the estimated cost:
• The number of informal workers is approximated using data on IMSS-insured workers
across these sectors. We assume that 58 percent of total workers across each sector are
informal (Alvarez and Ruane 2019). The total number of formal workers in construction,
trade, and services in 2019 was 10.5 million. If 58 percent of total employment consists
8
“Salaried workers are hired by entrepreneurs in relation of subordination and receive wages in return of their
efforts (fixed payments per unit of time)…..Non-salaried workers can be self-employed or associated with firms
but not in a subordinated relation, receiving remuneration in various forms (per unit produced or sold, per task
accomplished, profit-sharing, and the like), but not wages.” Levy (2018, pp.14).
9
According to Alvarez and Ruane (2019), informal firms include subsistence agriculture, domestic work, and
firms classified as informal by the Mexican Institute of Statistics and Geography (INEGI) based on reported
name, family ownership, and accounting practices. All other firms are formal.
15
of informal workers, this would imply there are 14.6 million informal workers.10 A
sector-specific lockdown parameter of around 0.7 is assumed, that is, around 70 percent
of workers in these industries might need help. This implies that around 10 million
informal workers from these sectors would require support.
• 3-month income support of MXN 2036.12 per month (average of urban/rural poverty and
extreme poverty lines, as computed by CONEVAL for July 2020) is provided to these 10
million informal workers. This would amount to MXN 61.6 billion, which is equivalent
to 0.3 percent of GDP.
The estimated cost is sensitive to the Figure 8. Cash Transfers to Informal Workers
assumptions of the amount and (Share of GDP)
duration of support and lockdowns.
0.6
To illustrate this caveat, Figure 8
plots the estimated cost of 0.5
supporting the informal workers for 0.4
these targeted sectors, varying the 0.3
lockdown assumptions and the 0.2
amount of monthly support. The
0.1
estimated cost ranges from 0.1-0.5
0.0
percent of GDP. The cost would 0.2 0.4 0.6 0.8 1
also depend on the duraton of the Lockdown (proportion of informal workers needing support)
support. The calcuation assumes 3 1500 pesos/month Poverty line (2036 pesos/month)
Minimum wage (2608 pesos/month)
months of support. However, if the
effects of the pandemic are prolonged—both in terms of the duration of the health crisis and
the subsequent pace of recovery—the support may need to be extended for longer.
Cost estimates of support to firms
Estimate #1. Wage subsidies to formal workers of hard-hit sectors (construction, trade, and
services), assuming sector-specific lockdown assumptions and six months of 2036 pesos
payments, would require support of 0.4 percent of GDP. The underlying calculation follows
the method explained above for informal workers with the same caveats applied. In
particular, the choice of the sectors has been determined using the same considerations
related to developments in phases 1 and 2, as described previously. The policy of subsidized
wages should be flexible and adjusted over time to mitigate the unintended adverse
consequences for labor market reallocation responses to the Covid-19 shock (Barrero, Bloom
and Davis, 2020).
Estimate #2: The banking sector credit to SMEs amounts to 11 percent of GDP. Assuming an
average maturity of 3 years, about 1 percent of GDP of credit falls due each quarter. The
rollover risk could be mitigated through development banks directly extending loans (against
10
This is an approximation since sectoral distributions do not necessarily match the aggregate distribution.
Using this approach for all sectors, total employment across all sectors is 49 million. Actual data suggests
employment of 56 million people. As such, this calculation could be regarded as a lower bound.
16
collateral) for about 1-2 percent of GDP. Alternatively, development banks could share (or
backstop) the credit risk with commercial banks to facilitate rollover.
Cost estimate of accelerated/additional investment
The authorities have already allocated 0.2 percent of GDP to social spending including public
works, security and education in response to Covid-19. This could be increased to 0.4 percent
of GDP, including investing in maintenance and restarting good projects that were delayed,
to recuperate a fraction of the private investment lost. The 2020 WEO projection for private
fixed investment has decreased from 17.7 percent of GDP (pre-Covid-19) to 16.5 percent of
GDP in July 2020 (WEO) version, a decline of 1.2 percentage points. Assuming that a
quarter of the lost investment could be replaced by public investment, we arrive at a cost
estimate of around 0.4 percent of GDP.
III. MEDIUM –TERM FISCAL REFORMS
Considering that social spending is set to increase rapidly over time (as a result of the Covid-
19 pandemic and, even before that, as a key priority for the current government) and other
expenditure pressures are emerging in Mexico, raising revenues from their current
exceptionally low levels will be indispensable to finance social spending sustainably and
create space for a more forceful near-term response to Covid-19 and its economic fallout.
There is also a need to increase public investment permanently and enhance social safety nets
to increase the political/public acceptability of the needed tax reforms.
A. Increase Tax Capacity
Despite impressive recent gains, Mexico’s tax revenues are the lowest in OECD and lag
regional peers. In 2014, the government passed a tax reform that set limits to deductions and
exemptions, introduced new tax brackets, created taxes on dividends and gains, and
introduced an excise tax on sodas, junk food, pesticides and carbon-producing products. As a
result, federal tax revenue increased by almost 3 percentage points of GDP, which helped
offset a large decline in oil revenues. However,
at 13.2 percent of GDP in 2019, non-oil tax
revenues remain well below the OECD and
Latin America averages of 26 and 19 percent of
GDP, respectively. Importantly, revenue from
VAT has stagnated and remains much below
that of other countries. This section proposes a
reform that would deliver at least 3 percent of
GDP in additional tax revenues. It would center
on policy and administration actions that could
improve VAT performance, rationalize
inefficient and regressive income tax
expenditures, and widen the top personal
income tax bracket.
17
Significant Scope to Improve VAT Performance
Value-added tax (VAT) collection in Mexico is low compared to OECD countries and Latin
America. In 2019, net of refunds, it reached only 3.9 percent of GDP (significantly less than
the OECD average of 7.0 percent and the Latin America average of 7.3 percent). The C-
efficiency of VAT at 29 percent— the ratio of actual VAT collections (ABFD area in Figure
10) to the theoretical revenues under a perfectly enforced tax levied at the standard rate on all
final consumption without any exemptions (ACHE area in Figure 10)— is one of the lowest
among OECD countries and Latin America (that has an average C-efficiency of 50 percent).11
The low level of VAT C-efficiency in Mexico can be explained by a narrow base (owing to
tax expenditures and informality) and by high non-compliance. In general, there are three
main factors that can explain a low level of tax collection in a country: (i) low rates; (ii) a
narrow base; and/or (iii) non-compliance (IMF 2018). At 16 percent, the standard VAT rate
in Mexico is not significantly lower than the averages of OECD countries and Latin America.
The base of the VAT in Mexico is reduced mainly for two reasons: (a) tax expenditures on
final products (such as exemptions of education and housing as well as zero ratings of
domestic transactions). These amounted to 1.43 percent of GDP in 2019 (0.26+1.17 in Table
3); and (b) exclusions from the object of the tax (owing to informality, for example). At the
same time, the VAT compliance gap is relatively high: estimated by IMF (2018) at 45.8
percent of potential revenue in 2016 (or 2.41 percent of GDP). Non-compliance and tax
expenditures are related: exemptions and zero rating of domestic transactions, in addition to
eroding the tax base, make control difficult and thus facilitate evasion.
Figure 10. Components of the Tax Gap Table 3. Tax Expenditures (Percent of GDP)
2018 2019 2020
VAT 1.361 1.432 1.432
Zero ratings 1.106 1.167 1.167
Exemptions 0.255 0.265 0.265
PIT 0.967 0.952 0.896
Deductions 0.118 0.102 0.102
Exemptions 0.718 0.733 0.683
Special or sectoral schemes 0.128 0.116 0.110
Deferrals 0.004 0.001 0.001
CIT 0.518 0.495 0.495
Deductions 0.089 0.099 0.099
Exemptions 0.051 0.047 0.047
Special or sectoral schemes 0.048 0.047 0.047
Deferrals 0.104 0.082 0.082
Administrative facilities 0.022 0.022 0.022
Employment allowances 0.205 0.198 0.198
Special Taxes 0.255 0.265 0.265
Exemptions 0.255 0.265 0.265
Fiscal Incentives 0.809 0.843 0.499
Source: IMF (2017). Source: SHCP.
11
Observed C-efficiency should equal the inverse of the total VAT gap, i.e., (1 - tax gap) or (1 - policy gap) x (1
– compliance gap). The compliance gap is the difference between the potential VAT that could have been
collected given the current policy framework and actual accrued VAT collections (BCGF in Figure 1). The
policy gap is the difference between the overall tax gap and the compliance gap.
18
Policy recommendations:
• Reduce policy gap. The “first-best” option to broaden the VAT base in Mexico includes
eliminating all domestic zero-rate items, except for a few key foodstuffs, and reducing
exemptions. These actions alone
could boost revenues by more
than 1 percent of GDP. They
should be accompanied by
spending measures to address
distributional concerns (through
strengthened social safety nets)
as, in general, VAT exemptions
and reduced rates are a poor tool
for protecting lower-income
households or improving income
distribution, given weak
targeting. While lower-income
households spend a greater share
of their income on consumption,
higher-income households consume absolutely more. This is shown in tax expense
estimates of the Mexican Finance Ministry (SHCP) for the year 2019, where the highest
income decile received 13.8 percent of the benefits from zero rate VAT on food, while
the lowest income decile received 4.1 percent of the benefits. Additional spending
through social safety net programs is a better way to improve equity. Among OECD
countries, Denmark, New Zealand, Chile and Estonia have opted for a broad-based,
simpler and more economically efficient VAT and redistribute part of the additional
revenue obtained through well-targeted social safety net programs.
• Reduce compliance gap. Mexico already has an adequate toolkit to combat tax evasion.
Staff recommends adopting a comprehensive strategy to tackle non-compliance in line
with the 2018 IMF technical assistance, which could result in revenue gains of 1 percent
of GDP. This includes improving the tax agency’s (SAT) fragmented organizational
structure, simplifying the small taxpayer regime, and moving toward a high-coverage
audit process for VAT returns. The recent abolition of the right to offset excess tax
credits against other taxes should reduce fraud once the backlog of grandfathered claims
clears. Strengthening sanctions against tax fraud would also be welcome.
Overall, fundamental VAT reforms to reduce both policy and compliance gaps over the
medium term could increase revenues by at least 2 percent of GDP (IMF 2018). If Mexico
were to raise the C-efficiency of VAT to the average of Latin America (50 percent), revenue
would increase by more than 2 percent of GDP. Such a revenue gain is much-needed, given
the expected marked decline in oil revenues over time and increased desire/need for
additional social protection and public investment. However, it has repeatedly proved
politically difficult since early 2000s. Moreover, as revenue gains from strengthened
compliance will take time to materialize, fiscal planning should be prudent and not ex ante
count on large gains (as assumes in model simulations). Accompanying measures to increase
acceptability by strengthening social safety nets could cost about 0.5 percent of GDP.
19
It should be noted that if “second-best” alternatives are pursued owing to political economy
constraints, it would be advisable to seek not only a partial improvement over the status quo
but also a path that ultimately approaches the “first-best.” For example, VAT exempting all
domestic zero-rated items would decrease the complexity of the tax structure and reduce the
number of VAT refunds, bringing some relief to the tax administration. However, the trade-
off would be limited overall revenue gains, with disproportionate leakage of benefits to the
better off and non-use of better means to support to the poor.
Some Scope to Increase PIT Revenues
The personal income tax (PIT) system is
performing very weakly in international
comparison. PIT revenues in Mexico are some
5 percent of GDP less than the OECD average,
reflecting informality and inequality with a
large low-wage sector paying very little tax,
given the progressivity of PIT. While there is
limited scope for increasing PIT revenue
through rate increases, the tax base can be
expanded. Exclusions from the PIT net are
prevalent such as income on personal business
activities and independent services. Apart from withholdings on wages and salaries, personal
income tax revenues are negligible. Tax expenditures for PIT are also sizable: close to 1
percent of GDP in 2019. The authorities consider that at least 0.5 percent of GDP of these tax
expenditures are inefficient or regressive and could be rationalized (IMF 2019c). Moreover,
the threshold for the top PIT bracket (which has been increased by five percentage points as
part of the 2014 tax reform) could be lowered to widen the base, as it is a significant outlier
compared to peers. Overall, the potential revenue gains from PIT reforms are about 0.5-1
percent of GDP.
Figure 13. Progressivity of PIT Figure 14. Bottom and Top PIT Brackets
(Percent and USD) (Percent of per-capita GDP)
Sources: IMF staff estimates using IBFD data. Source: FAD Revenue Analysis Tool.
Note: Lines show the Average Effective Tax Rate (AETR) by
Personal Income Level.
20
Limited Scope to Increase CIT Revenue
Mexico’s corporate income tax (CIT) system features a relatively high tax rate and revenue,
and a fairly typical base. The CIT rate is a flat 30 percent, which is above the U.S. federal
rate of 21 percent and the average of OECD countries. CIT revenue makes up a quarter of the
low overall tax take and, at 3.6 percent of GDP, compares favorably with the OECD average
of 2.8 percent. Such a strong reliance on CIT revenue as well as rate differentials expose
Mexico to potential revenue losses associated with profit-shifting and, over the medium term,
relocation of production if rates were to be increased (which however is unlikely).12 The tax
base is also fairly typical and tax expenditures under the CIT are limited (0.5 percent of
GDP) and likely overstated. This is because deferred taxes, a part of deductions and
administrative facilities reflecting genuine business costs, and employment subsidies (partly
addressing the unusual feature of the PIT system, where income is taxed from the first peso)
are included in tax expenditures. Overall, considering that the productivity of CIT is at the
average of international peers and above that of several advanced economies, there seems to
be little scope to increase CIT revenue through base broadening measures. The revenue
agency (SAT) has been diligently pursuing collections against several corporations in recent
months, including in a number of high-profile settlements. The above-mentioned assessment
of CIT taxes, however, notes that significant further scope for revenue collection in this area
is not likely to be as much as similar efforts, alongside tax policy reform, in the areas of VAT
and PIT.
Figure 15. Corporate Income Tax Rates and Productivity
12
An important feature of the Mexican tax system is the treatment of certain export-oriented operations, known
as the maquila sector, which are taxed under safe harbor or Advance Pricing Agreement rules, implying that
their CIT base is estimated as a percentage of their assets or costs.
21
Other taxes
Gasoline excise tax. The current formula guarantees cumulative retail fuel price growth
below CPI inflation since November 30, 2018. This policy disproportionately benefits the
rich and should be revoked, which could provide some ¼ percent of GDP in additional
revenues. Moreover, in a low oil price environment, there is scope to increase gasoline
excises.
Subnational taxes. Over the medium term, there is much potential to increase revenue from
strengthening property taxation—which is
currently 1.5 percent of GDP less than the average
of Latin American countries. A reform could be
facilitated by creating an agency at the federal
level to update the cadaster, as well as by policy
coordination at the subnational level to avoid
fragmentation. Closing the gap in collections vis-
à-vis peers could take time but, at a minimum, the
aim could be to collect at least ¾ percent of GDP.
This could be complemented with a redesigned
vehicle registration tax by simplification and
stricter enforcement, hence enabling a reduction in
transfers to states and municipalities.
B. Raise Public Investment
In the medium term, the proposed package includes an increase in public investment of 1.5
percent of GDP, out of which a third is allocated to healthcare.
Public investment. High-return public
investment can act as a bridge to sustainable, Figure 17. Public Gross Fixed Capital Formation
(Percent of GDP)
resilient, and inclusive economic growth, 10
including by lifting productivity, creating jobs, 9
8
and spurring private sector investment (IMF 7
2020b). In the case of Mexico, IMF (2019a) 6
5
shows that there is a crucial role for basic 4
infrastructure investment (in particular, road 3
2
infrastructure) in boosting firm productivity, 1
including for the large number of micro firms. 0
In this regard, the trend decline in public
Mexico LA-6 EMs OECD
investment-to-GDP ratios in Mexico is Source: WEO estimates.
LA-6: Argentina, Brasil, Chile, Colombia, Mexico and Peru.
concerning. Investment spending, as a share of EMs: ARG, BRA, CHL, CHN, COL, EGY, HUN, IND, IDN. MYS, MEX, PER, PHL, POL, ROM, RUS,
ZAF, THA, TUN, TUR.
GDP, has declined from 4 percent of GDP over
2008-2018, on average, to 2½ percent in 2019. Capital spending in Mexico is also lower than
regional and EM peers and OECD countries. . Consequently, the proposed package includes
an increase of public investment of around 1 percent of GDP in the medium-term (excluding
investing in healthcare; see below) to bring the ratio in line with historical averages and that
of peers.
22
Health spending. As computed in IMF (2019a), an increase in health spending of ½ percent
of GDP is needed over the medium-term to make satisfactory progress towards the
Sustainable Development Goals (SDGs) in health by 2030. The costing exercise follows the
input-output methodology of Gaspar and others (2019) in: (i) identifying key inputs and their
corresponding costs in the health sector; (ii) benchmarking input costs in Mexico to that of
countries with comparable income per capita but higher current social outcomes; and (iii)
estimating the spending levels required to achieve those outcomes, conditional on Mexico’s
income per capita and population growth. IMF (2019a) finds that Mexico could increase the
share of doctors in the population and their wages but contain the number of other health
professionals. It must also be noted that the spending to achieve SDG goals would be
considerably larger if efficiency gains are not achieved. This is probably a conservative
estimate given the likely long-lasting effects of the Covid-19 shock.
C. Strengthen Social Safety Nets and Unemployment Insurance
In response to the Covid-19 shock, the need for emergency lifelines would be sizable, albeit
lower—all else being equal—for countries with stronger unemployment benefits and social
safety nets (SSN).13 Given the lack of a nationwide unemployment benefits system in Mexico
and the likely adverse impact of the pandemic on poverty and inequality for years to come,
there is a need to strengthen SSNs durably. An enhanced SSN can also protect the most
vulnerable from unintended consequences of proposed VAT reform.
A good social safety net usually has four attributes (Grosh and others 2008 and IMF 2020b).
First, it provides broad coverage and adequate benefits to vulnerable groups in a progressive
way within the overall tax-benefit system (IMF 2019b)—that is, more generous benefits to
accruing to the poorest beneficiaries. Second, it strives to be cost effective by avoiding
program fragmentation and beneficiary overlaps. Third, it aims to preserve work incentives
and enhance human capital accumulation by linking transfers to required or voluntary
programs such as public works, obtaining health care, and attending education and training.
Fourth, it is financially sustainable within the overall expenditure envelope and consistent
with other social protection programs.
Against these yardsticks, SSNs in Mexico have significant gaps in terms of coverage of
lower income groups and beneficiary incidence—see Figures 18 and 19. The cost of SSNs as
a share of GDP is relatively high, in part, reflecting social and political choices made to
reduce poverty gaps (Figures 20 and 21). Programs are fragmented, involve beneficiary
overlaps, and lack appropriate incentive features (IMF 2019a). Moreover, the burden of
income support is placed on SSNs, as very few of the poor are covered by unemployment
insurance. SSNs in Mexico can be improved by expanding coverage in a cost-effective
manner through proxy-means targeted transfers, whereby targeting is improved by giving
13
Further to the discussion in footnote 6, social safety nets are noncontributory transfer programs aimed at low-
income households or the vulnerable (World Bank 2018; IMF 2020b). They are financed through government
revenues and usually comprise (i) cash transfers, food stamps, child allowances, and social pensions; (ii) in-kind
transfers; (iii) conditional transfers; (iv) public works; and (v) fee waivers, including for health care.
23
households a score based on a statistical algorithm that predicts incomes or consumption or
poverty (see Coady and Le 2020 for details), and decreasing duplications as the country
strengthens its administrative capacity. The relatively large leakage of benefits to higher-
income groups in Mexico increases the importance of better targeting as well as
strengthening progressive income taxes to claw back these benefits from high-income
groups. Mexico can also use instruments that are effective in reaching individuals most in
need, including in the informal sector. These instruments include mobile money, in-kind
transfers such as education and health, matching different databases of beneficiaries to create
a single registry, and use of community-based methods to identify those in need. Beyond
SSNs, there is scope in designing a nation-wide unemployment benefits system.
Figure 18. Coverage and Adequacy of Social Safety Net Programs by Quintile
(Percent)
1. Coverage 2. Adequacy
Source: IMF FAD Social Protection & Labor - Assessment Tool (SPL-AT).
Notes: Coverage measures the percentage of the quintile that receives a SSN benefit. Adequacy of benefits measures the
total transfer amount received by all beneficiaries in the quintile as a percent of the pre-transfer total income/expenditure
of beneficiaries (in that quintile). Calculations are based on pre-Covid-19 information.
Figure 19. Targeting of Social Safety Net Programs by Quintile
(Percent)
1. Benefits Incidence 2. Beneficiary Incidence
Source: IMF FAD Social Protection & Labor - Assessment Tool (SPL-AT).
Notes: Benefits incidence measures the percentage of benefits going to each quintile of the pre-transfer welfare
distribution relative to the total benefits going to the population. Beneficiary incidence measures the percentage of
program beneficiaries in a quintile relative to the total number of beneficiaries in the population. The indicator is
estimated by program type and by quintiles of the pre-transfer welfare distribution. Calculations are based on pre-Covid-
19 information.
24
Figure 20. Composition of SSNs Figure 21. Poverty Reduction Impact of
(Percent of GDP) SSNs (Percent of GDP)
Source: ASPIRE (World Bank) and national authorities. Source: IMF World Economic Outlook.
Note: Apart from Mexico, which reports the size of its SSN Note: Simulated percentage change reduction in poverty gap
in 2020, all the other numbers are the latest available owing to social safety net programs. The calculation is: (poverty
before the outbreak of Covid-19. gap pre transfer- poverty gap post transfer) / poverty gap pre-
transfer.
IV. MACROECONOMIC IMPACT OF THE PROPOSED PACKAGE
This section uses the IMF’s Global Integrated Monetary and Fiscal (GIMF) model to
illustrate the macroeconomic effects of a “fiscal twist” strategy for Mexico.14 GIMF is a
multi-country structural dynamic general equilibrium model featuring Mexico, among others.
It links the behavior of households, firms, and government sector within and among countries
(see Box 2 for details), and is particularly suitable for fiscal policy analysis.
A. The “fiscal twist” package
Informed by the analysis in Sections II and III, a comprehensive fiscal policy response to
Covid-19 and beyond is designed along the following dimensions (summarized in Table 4):
• In the first year, stronger emergency lifelines in response to Covid-19 is assumed:
o Expenditure. Increase of 0.5 percent of GDP in government consumption (largely
healthcare); 0.5 percent of GDP in government investment (out of which 0.1 percent
of GDP is on healthcare); and 1.2 percent of GDP in targeted social transfers.
o Revenue. 0.3 percent of GDP in wage subsidies which is reflected in the model as a
decline in labor tax.
• Broad-based fiscal support in the second year to facilitate the recovery followed by
growth-friendly and inclusive adjustments thereafter (fully anticipated):
14
With the economy re-opening and signs of a recovery underway, use of a standard framework (GIMF) can
help unpack the benefits and costs of the proposed package.
25
o Expenditure. Maintain higher levels of public investment to support the recovery
next year and over the medium-term for healthcare (0.5 percent of GDP; to make
progress toward the Sustainable Development Goals in health). Enhancing social
safety nets would cost about 0.5 percent of GDP. They should support the
government’s poverty-reduction efforts (0.2 percent of GDP) and alleviate the
regressive nature of the proposed VAT reform (0.3 percent of GDP)..
o Revenue. As a result of the proposed tax reform, revenues would structurally
increase by 2.0 percent of GDP from the third year, assuming the pandemic is fully
under control by then. This includes broadening of the PIT base and increasing
social security contributions (0.5 percent of GDP), improving VAT performance (1
percent of GDP), broadening of the CIT base (0.2 percent of GDP), and increasing
property/local taxes (0.3 percent of GDP). Revenue gains are assumed to be lower
than what we identified in Section III in the simulations. The remainder is assumed
to close estimated fiscal gaps (in the baseline).
Overall, the above package assumes a 2½ percent of GDP increase in the overall fiscal deficit
in the first year to mitigate the economic fallout from Covid-19 through emergency lifelines.
This fiscal support is tapered off in the second year to 1 percent of GDP and its focus is
changed from lifelines to facilitating the recovery through long-lasting public investment and
stronger social safety nets. A 2 percent of GDP permanent increase in taxes (VAT, corporate
and personal income taxes, and lumpsum taxes) is assumed in the third year alongside an
additional increase in public investment (1 percent of GDP). Overall, this would result in a
neutral-budget package from the third year as the increase in revenues offsets the increase in
higher-quality expenditure. This neutral package is assumed above-and-beyond measures to
close fiscal gaps that remain in the baseline. Assuming that the medium-term tax reforms are
perceived as credible, the package would lower the risk premium by 50 basis points owing to
positive confidence effects from the much-needed reforms.
Finally, a scenario is simulated where public investment efficiency is assumed to be 30
percent higher. On average, more than one-third of funds for public infrastructure are
estimated to be lost owing to inefficiencies worldwide (IMF 2015; Baum, Mogues, and
Verdier 2020). Mexico is no exception to this problem and to increase the long-term output
gains from increased public investment, efficiency needs to be improved. Sound institutional
processes, including careful project selection, management, and evaluation, as well as a clear
delineation of responsibilities and mechanisms to ensure coordination between central and
subnational governments, should be in place to ensure productive investment (IMF 2015).
26
Table 4. Model Inputs for GIMF Simulation
(In percent of January 2020 WEO GDP)
t=1 2 3 4 5 6
Government deficit (expenditure-revenue) 2.5 1.0 0.0 0.0 0.0 0.0
Total expenditure 2.2 1.0 2.0 2.0 2.0 2.0
Government consumption 1/ 0.5
Government investment 0.5 0.5 1.5 1.5 1.5 1.5
Targeted transfers 1.2 0.5 0.5 0.5 0.5 0.5
Total revenue -0.3 0.0 2.0 2.0 2.0 2.0
Labor tax (incl. social security contribution) -0.3 0.5 0.5 0.5 0.5
Consumption tax 1 1 1 1
Corporate Income tax 0.2 0.2 0.2 0.2
Property tax (lumpsum tax) 0.3 0.3 0.3 0.3
Sovereign risk premium (basis point) -50 -50 -50 -50 -50 -50
1/ Health spending classified as part government consumption, part government invesment.
Box 2. The Global Integrated Monetary and Fiscal Model (GIMF)
The IMF’s Global Integrated Monetary and Fiscal (GIMF) model is used to quantify the
effects of the fiscal package—see Kumhof and others (2010) and Anderson and others (2013)
for more details.
GIMF is a multi-country structural dynamic general equilibrium model featuring six regions:
Mexico, the US, Japan, emerging Asia, euro area, and the rest of the world. It links the
behavior of households, firms, and government sector within and among countries. The model
has a consistent system of national accounting and stock-flow budget constraints for all
sectors, including the government. The model belongs to exogenous-growth types of models,
meaning that the long-term growth of output is exogenous. Hence, all fiscal or structural
measures may change only the structure of the economy, possibly increasing permanently the
level of real output per capita, and not the long-term growth.
The household sector consists of forward-looking optimizing households, as well as liquidity-
constrained households who spend all their available income every period. The forward-
looking households are modeled as overlapping generations (OLG) with finite lives, following
the Blanchard-Weil-Yaari framework. The presence of OLG households breaks Ricardian
equivalence and is important for realistic results of fiscal policy in the short and long runs.
Households gain utility from consumption and disutility from labor effort, consume traded
and non-traded services and goods, receive labor income, transfers from the government and
dividends from corporations, and pay income, consumption and lump-sum taxes.
Firms produce intermediate and final goods using labor and capital inputs and accumulate
capital. Firms pay taxes from corporate income.
27
The government collects tax revenues (consumption, labor income, capital income, and lump-
sum taxes) and spends them on government consumption, investment, and lump-sum transfers
to either all or targeted to liquidity-constrained households. It targets a specific debt-to-GDP
(and thus deficit-to-GDP) ratio and uses a mix of instruments to achieve it. The government’s
commitment to sustainable public finance is credible for firms and households, who hold the
stock of government bonds.
Fiscal policy is conducted using a variety of expenditure and tax instruments. Beside the
afore-mentioned instruments, the model also allows for tariffs on imported goods to be a
potential source of public revenue. Government investment spending augments public
infrastructure, which depreciates at a constant rate over time.
There is a fiscal policy rule that ensures long-run sustainability, while allowing for short-run
counter-cyclical policies. Changes in labor and capital income taxes provide the instruments to
put the rule into effect, but this can be replaced with other tax, transfer or spending
instruments if that is considered more realistic for a specific region. First, the fiscal rule
ensures that, in the long run, the ratio of the government debt-to-GDP—and hence the deficit-
to-GDP ratio—eventually converges to its target level. This excludes the possibility of
sovereign default, as well as the risk that out-of-control financing requirements of the
government will override monetary policy. Second, the rule allows for countercyclical fiscal
policy as it embodies automatic stabilizers.
When conducting monetary policy, the central bank uses an inflation-forecast-based interest
rate rule. The central bank varies the gap between the actual policy rate and the long-run
equilibrium rate to achieve a stable target rate of inflation over time.
Calibration
The calibration of the parameters of the Mexico: Key Calibration Parameters
household utility function, elasticities and (Percent)
markups are the same across countries. The Inflation 3.0
shares of liquidity-constrained agents in the Share of liquidity constraned households 50
population are 50 percent for Mexico, Government consumption/GDP 15.7
Emerging Asia and the remaining countries Net acquisition of nonfinancial assets/GDP 1.3
and 25 percent for other regions. Country Private investment/GDP 18.6
specific data are used for great ratios based Government transfer/GDP 5.4
on the most recent year or some average Tax and SSCs revenue/GDP 15.4
(all in percent of GDP). Key variables Labor tax incl. SSCs/GDP) 5.6
include private investment share, Consumption tax (% GDP) 5.8
government consumption, public CIT revenue (% GDP) 3.6
investment, fiscal variables including tax Property and other taxes/GDP 0.5
revenue, government debt, share of Other revenue/GDP 8.7
different taxes in GDP (consumption tax, Gross public debt/GDP 53.7
labor and corporate income tax), and labor
shares. In addition, bilateral trade flows between regions and each region’s share of world
population are essential in accounting for spillover effects. The monetary policy parameters
are calibrated based on estimation results of reaction functions for an annual model.
28
B. Model Results
We present the macroeconomic effects
Figure 22. Macroeconomic Effects of the Fiscal Package
of the “fiscal twist” package compared
to the January 2020 WEO baseline in Real GDP
(Percent deviation from baseline)
three blocks: (i) expenditure measures; 6
(ii) the total package, including the Total with monetary accommodation
5
expenditures measures, tax reforms, + Public investment efficiency
4
and 50 bps decline in risk premia; and
(iii) the total impact assuming 3
continued monetary accommodation 2
so long as inflation expectations
1
remain anchored and demand
0
shortages persist.15 Furthermore, the t=0 1 2 3 4 5 6 t=11
results with the increase in public
Government Deficit/GDP
investment efficiency are presented (Deviation from baseline)
(Figure 23). Note that the Covid-19 2.5 Total with monetary accommodation
shock has already resulted in a sharp 2.0 + Public investment efficiency
GDP contraction in 2020 with 1.5
expectation of a gradual economic 1.0
recovery over the medium term. The 0.5
payoffs from the “fiscal twist” 0.0
package (presented below) would, -0.5
therefore, only be able to restore a part
-1.0
of the GDP fall. t=0 1 2 3 4 5 6 t=11
Government Net Debt/GDP
The expenditure measures are (Deviation from baseline)
2.0
expected to increase the level of GDP Total with monetary accommodation
by 1.2 percent in the first year, and, 1.0
+ Public investment efficiency
over the medium-term, by 2 percent. 0.0
Increased demand within the economy
-1.0
is expected to appreciate the real
-2.0
exchange rate and deteriorate the
current account balance. The overall -3.0
budget deficit would increase -4.0
according to the size of spending t=0 1 2 3 4 5 6 t=11
increases in Table 4. Without any
concomitant reforms to boost
revenues, public debt would increase by 9 percentage points of GDP in the medium term.
Combining the expenditure measures with a credible tax reform (including risk premia
reductions) would lead to output gains as well as a stable public debt path (assuming no
additional monetary accommodation). The level of GDP would be higher relative to the
15
Estimates from the empirical literature inform the multipliers assumptions in GIMF. See IMF 2020c for
details on multiplier estimates by instrument.
29
January 2020 WEO baseline in the first year as a result of larger emergency lifelines.
Following a dip in the second year as fiscal support is partially withdrawn, output is expected
to continue to increase by about 3.6 percent in the long run (in eleven years). The real
exchange rate is expected to appreciate in the first few years owing to the increased domestic
demand, but eventually depreciate as the tax reforms are enacted. The effect of domestic
demand and the real exchange rate would deteriorate current account initially but then
improve over the medium term. The government deficit will increase in the first year and
then gradually decrease.
The overall positive effect on GDP and public debt trajectory is greater when monetary
authorities do not react to the higher fiscal deficits by raising interest rates so long as
inflation expectations remain well anchored and demand shortages persist (i.e., the Central
Bank could remain accommodative within its mandate). In the medium term, the output
could increase by 3.2 percent, government deficit could fall 0.4 percentage points, and
government debt could fall by 2.2 percentage points. Finally, the increase in investment
efficiency would be further beneficial, with output improving by 4.1 percent, government
deficit falling by 0.6 percentage points, and government debt declining by 3.5 percentage
points over the medium term. The gains from improvements in the composition of fiscal
policy mix and the fiscal twist are sizable. Coupled with structural reforms to improve
productivity growth, the overall benefits would be very significant over the medium term.
30
Figure 23. Impact of Proposed Fiscal Package
3
Real GDP (percent deviation from baseline) Real Consumption (percent deviation from baseline) 4
4.5 4.0 5
4.0 6
3.5
7
3.5
3.0 8
3.0
2.5
2.5
2.0
2.0
1.5
1.5
1.0 1.0
0.5 0.5
0.0 0.0
t=0 1 2 3 4 5 6 t=11 t=0 1 2 3 4 5 6 t=11
Real Investment (percent deviation from baseline) Current Account/GDP (deviation from baseline)
8.0 0.0
7.0
-0.5
6.0
5.0 -1.0
4.0
3.0 -1.5
2.0
-2.0
1.0
0.0 -2.5
t=0 1 2 3 4 5 6 t=11 t=0 1 2 3 4 5 6 t=11
Real Effective Exchange Rate Interest Payments/GDP (difference from baseline)
1.5
(percent deviation from baseline; +ve=appreciation) 1.2
1.0 1.0
0.5 0.8
0.0 0.6
-0.5 0.4
-1.0 0.2
-1.5 0.0
-2.0 -0.2
-2.5 -0.4
t=0 1 2 3 4 5 6 t=11 t=0 1 2 3 4 5 6 t=11
Government Deficit/GDP (difference from baseline) Government Net Debt/GDP
3.0 16.0 (difference from baseline)
2.5 14.0
12.0
2.0
10.0
1.5 8.0
1.0 6.0
0.5 4.0
2.0
0.0
0.0
-0.5 -2.0
-1.0 -4.0
t=0 1 2 3 4 5 6 t=11 t=0 1 2 3 4 5 6 t=11
31
V. CONCLUSION
Mexico needs to contain the pandemic, mitigate its economic effects, facilitate a rapid
recovery, and rebuild fiscal buffers over the medium term. Fiscal policy is an important lever
to meet these goals. Fiscal resources are needed in the near term to accommodate additional
spending on healthcare and emergency services as well as to protect people and viable firms
through timely, targeted and temporary lifelines. In the near term, the paper estimates that a
fiscal support of around 2.5-3.5 percent of GDP would be required in health spending, social
safety nets, wage subsidies, and public investment.
Given that social spending are projected to increase over time and other expenditure
pressures are emerging in Mexico (including the need for higher public investment), raising
revenues from their current exceptionally low levels is indispensable to finance them
sustainably. A higher revenue envelope could also create fiscal space for a more forceful
near-term Covid-19 response. The paper proposes a tax reform of at least 3 percent of GDP,
centered on policy and administration actions that could improve VAT performance,
rationalize inefficient and regressive income tax expenditures, and widen the top personal
income tax bracket. This reform should be paired with enhanced social safety nets.
Using IMF’s GIMF model, simulations suggest that the proposed “fiscal twist” package
would substantially increase output in the medium term and put public debt on a firm
declining path compared to their January 2020 WEO path. Higher investment efficiency
would further bolster these effects. While the focus of the paper is on fiscal policies,
supportive stances in monetary and other policies would be required for a durable economic
recovery.
32
VI. REFERENCES
Alvarez, J. and C. Ruane, 2019, “Informality and Aggregate Productivity: The Case of
Mexico,” IMF Working Paper WP/19/257 (Washington: International Monetary Fund).
Anderson, D., B. Hunt, M. Kortelainen, M. Kumhof, D. Laxton, D. Muir, S. Mursula, and
S. Snudden, 2013, “Informality and Aggregate Productivity: The Case of Mexico
Getting to Know GIMF: The Simulation Properties of the Global Integrated Monetary
and Fiscal Model,” IMF Working Paper WP/13/55 (Washington: International
Monetary Fund).
Barrero, Jose Maria, Nicholas Bloom, and Steven J. Davis. 2020. “COVID-19 Is Also a
Reallocation Shock.” NBER Working Paper 27137, National Bureau of Economic
Research, Cambridge, MA.
Baum, Anja, Tewodaj Mogues, and Genevieve Verdier. 2020. "Getting the Most from Public
Investment." In Infrastructure Governance: From Aspiration to Action, edited by
Gerd Schwartz, Manal Fouad, Torben Steen Hansen, and Genevieve Verdier.
Washington, DC: International Monetary Fund.
Chudik, Alexander, Kamiar Mohaddes, M. Hashem Pesaran, Mehdi Raissi, and Alessandro
Rebucci. 2020. “A Counterfactual Economic Analysis of Covid-19 Using a Threshold
Augmented Multi-Country Model.” NBER Working Paper 27855, National Bureau of
Economic Research, Cambridge, MA.
Chudik, Alexander, Kamiar Mohaddes, M. Hashem Pesaran, and Mehdi Raissi. 2017. "Is
There a Debt-Threshold Effect on Output Growth?" The Review of Economics and
Statistics 99 (1): 135–50.
Coady, D. and N. Le, 2020. “Designing Fiscal Redistribution: The Role of Universal and
Targeted Transfers”, IMF Working Paper WP/20/105 (Washington: International
Monetary Fund).
Deb, Pragyan, Davide Furceri, Jonathan D. Ostry, and Nour Tawk, 2020. “The Economic
Effects of COVID-19 Containments Measures”, IMF Working Paper WP/20/158
(Washington: International Monetary Fund).
Dudine, P., K. Hellwig, and S. Jahan, 2020. “A Framework for Estimating Health Spending
in Response to COVID-19”, IMF Working Paper Series, WP/20/145 (Washington:
International Monetary Fund).
Kumhof, M., D. Laxton, D. Muir and S. Mursula, 2010, “The Global Integrated Monetary
Fiscal Model (GIMF) —Theoretical Structure,” IMF Working Paper Series, WP/10/34
(Washington: International Monetary Fund).
Gaspar, V., D. Amaglobeli, M. Garcia-Escribano, D. Prady, and M. Soto, 2019. Fiscal Policy
and Development: Human, Social, and Physical Investments for the SDGs.
International Monetary Fund.
Grosh, Margaret E., Carlo Del Ninno, Emil Tesliuc, and Azedine Ouerghi. 2008. For
Protection and Promotion: The Design and Implementation of Effective Safety Nets.
Washington, DC: World Bank.
33
International Monetary Fund (IMF) 2015. "Making Public Investment More Efficient." IMF
Policy Paper, Washington, DC.
IMF 2017. “Mexico Revenue Administration Gap Analysis Program- The Value-Added Tax
Gap”, Technical Assistance Report, May 2017.
IMF 2018. “Mexico: Evaluación del Desempeño y Nivel de Cumplimiento del IVA”.
Reporte de Asistencia Técnica.
IMF 2019a. “Mexico: Budgetary Spending Pressures”, Mexico: Selected Issues, IMF
Country Report No. 19/337, International Monetary Fund.
IMF 2019b. "A Strategy for IMF Engagement on Social Spending." IMF Policy Paper
19/016, Washington, DC.
IMF 2019c. “Mexico: 2019 Article IV Consultation Staff Report”, IMF Country Report No.
19/336, International Monetary Fund.
IMF 2020a. “October 2020 Regional Economic Outlook: Western Hemisphere — Fiscal
Policy at the Time of a Pandemic: How has Latin America and the Caribbean Fared?”
International Monetary Fund, Washington, DC.
IMF 2020b. “Fiscal Monitor: Policies to Support People During the COVID-19 Pandemic.”
Fiscal Affairs Department, International Monetary Fund, Washington DC, April 2020.
IMF 2020c. “Managing the Impacts of the Coronavirus: Guidance on Health Spending
Policies.” Special Series on Fiscal Policies to Respond to COVID-19:
https://www.imf.org/~/media/Files/Publications/covid19-special-notes/enspecial-series-
on-covid19managing-the-impacts-of-the-coronavirus-guidance-on-health-spending-
polic.ashx?la=en.
IMF 2020d. “October 2020 Regional Economic Outlook: Western Hemisphere — Covid-19
in Latin America and the Caribbean—A High Toll on Lives and Livelihoods?”
International Monetary Fund, Washington, DC.
Levy, Santiago. 2018. Under-rewarded efforts: The Elusive Quest for Prosperity in Mexico.
Inter-American Development Bank. https://flagships.iadb.org/en/Under-Rewarded-
Efforts.
Parr 2020. “COVID-19 Data Analysis, Part 5: Different Models of Infection Rates in Mexico
and What they Tell Us.” https://dai-global-digital.com/covid-19-part-5-different-
methods-to-model-infection-rates-in-mexico-and-what-they-tell-us.html
World Bank 2018. The State of Social Safety Nets 2018. Washington, DC: World Bank.
Zavala, M. R. and L. Despeghel, 2020. “¿Qué nos dicen las actas de defunción de la CDMX?
Actualización al 23 de agosto 2020”, Nexos. https://datos.nexos.com.mx/?p=1657
File and source
- File
- wpiea2020215-print-pdf-ashx_1694d696731cfd17.pdf
- Size
- 1,371,004 bytes
- SHA-256
- 420e75046910ab256c265aa091ab3f868cdef683266266c49b74d57e6bc5eee9
- Original
- www.imf.org