Crisis and poverty: With reference to Latin America

Abstract

This study analyzes the relationship between four types of economic crises and four poverty indices in an effort to determine which type of crisis most affects the poor, and find possible solutions. This study is particularly concerned with Latin American countries in which International Monetary Fund bailout programs have failed due to repeated crises, longer lasting inflation, and most of all poverty. The results indicate an apparent worsening trend in poverty measures due to the crises. Among several types of crises, debt crisis and currency crisis play larger roles than others, particularly with the headcount poverty ratio US$3.10.

Keywords

Economic crisis poverty indicator Latin America panel data

Introduction

Crisis impedes economic activity, undermining the consumption levels and lowering the demand for investments. Depressed economic activity leads to lower growth rates, higher unemployment rates, and lower wages, which eventually has a negative impact on the individual standard of living. Do crises seriously affect a country’s poverty level? If so, which type of crisis plays major roles and how can the effects be mitigated? Hopkins (2006) examined the relationship between crisis and health in three countries – Indonesia, Malaysia, and Thailand – and found that economic crisis led to a decline in real household income and government tax revenue, which diminished access to high quality food, health facilities, and education. A number of other studies have examined a variety of countries – Suryahadi and Sumarto (2003), Dhanani and Islam (2002), and Manning (2000) on Indonesia; Lokshin and Ravallion (2000) on Russia; Ahmed and O’Donoghue (2010) on Pakistan, and Breisinger et al. (2011) on Yemen.

Latin America is of particular importance in this study; it has suffered from repeated crises (see Table 1 for the list of countries and the crises they had experienced), unstable macroeconomic conditions (see Table 2 for government debt stock), long lasting hyper-inflation (see Table 3 for inflation and consumer price), and unsuccessful domestic monetary and fiscal policies, as well as unsuccessful bailout packages from the International Monetary Fund (IMF). Moreover, the region is closely associated with the United States – geographically, politically, and economically – which makes the region vulnerable to shocks stemming from the world’s largest economy. Against this backdrop, several studies have investigated the impact of crisis on Latin American countries. Ocampo (2009) focused on the impact of “transmission of crisis” – remittance, trade, and banking – on several economic factors and found that all 18 Latin American countries had experienced a decrease in current account balance, with the biggest drop in Chile, up to 7.4%, during the 2007–2008 period. Nikoloski (2011) examined the impact of currency, banking, and debt crisis on poverty measures and found the currency crisis is related with a higher “level” of poverty while the banking crisis explains the higher “depth” of poverty. Ocampo (2009) examined the theoretical connection between economic crisis and the poverty indicator, and Laeven and Valencia (2008, 2012) provided empirical evidences for selecting explanatory variables in terms of the government sector, external factors, and economic conditions.

Table 1.

Latin American countries experiencing global crisis (2006–2013).

Country	Code	Year
Country	Code	2006	2007	2008	2009	2010	2011	2012	2013
Argentina	ARG	0	0	0	0	0	0	1	1
Bolivia	BOL	0	0	1	0	0	1	0	1
Brazil	BRA	0	0	0	1	1	1	0	0
Chile	CHL	0	0	1	0	0	0	0	0
Colombia	COL	0	0	0	0	0	0	0	0
Costa Rica	CRI	0	0	1	0	0	0	1	1
Ecuador	ECU	0	0	1	0	0	0	0	0
El Salvador	SLV	0	0	1	0	0	1	1	1
Honduras	HND	0	0	0	0	1	0	0	0
Jamaica	JAM	0	0	0	1	1	0	0	0
Mexico	MEX	0	0	1	1	1	1	1	1
Nicaragua	NIC	0	1	1	0	0	0	0	0
Panama	PAN	0	1	1	0	0	1	1	1
Paraguay	PRY	0	0	0	0	0	0	0	0
Peru	PER	0	0	1	0	0	0	0	0
Uruguay	URY	0	0	1	0	0	0	0	0
Venezuela, RB	VEN	0	0	0	1	1	1	1	1

On top of these, the following shows other types of crises in the earlier period (Laeven and Valencia (2008, 2012) and Ocampo (2009)). Argentina: banking crisis (1990–1991, 1995, 2001–2003), currency crisis (2002–2004), debt crisis (2001–2002), banking crisis (1994). Brazil: banking crisis (1990–1998), currency crisis (1999–1901). Colombia: banking crisis (1998–2000). Costa Rica: banking crisis (1994–1995), currency crisis (1990–1991). Ecuador: banking crisis (1998–2002), currency crisis (1999–2001), debt crisis (1999–2000). El Salvador: banking crisis (1990). Honduras: currency crisis (1990–1992). Jamaica: banking crisis (1996–1998), currency crisis (1991–1993). Mexico: banking crisis (1994–1996), currency crisis (1995–1997). Nicaragua: banking crisis (1990–1993, 2000–2001), currency crisis (1990–1992). Paraguay: banking crisis (1995), currency crisis (2002–2004). Uruguay: banking crisis (2002–2005), currency crisis (2002–2004), debt crisis (2002–2003). Venezuela, RB: banking crisis (1994–1998), currency crisis (1994–1997, 2002–2004).

Table 2.

Central government public debt stock as percentage of GDP for 17 Latin American countries.

Country	Code	Year								Mean	Median	SD
Country	Code	2006	2007	2008	2009	2010	2011	2012	2013	Mean	Median	SD
Argentina	ARG	70.6	62.1	53.8	55.4	43.5	38.9	40.4	43.5	51.025	48.65	11.376
Bolivia	BOL	49.7	37.2	34	36.3	34.6	34.5	29.1	28.4	35.475	34.55	6.553
Brazil	BRA	56	57	57	60	52	51	55	56.7	55.5875	56.35	2.906
Chile	CHL	5	4	5	6	9	11	11	13	8	7.5	3.423
Colombia	COL	40	37	36	38	39	37	33	37	37.125	37	2.100
Costa Rica	CRI	32	27	24	27	28	30	34	36	29.75	29	4.027
Ecuador	ECU	26	25	21	15	18	17	20	23	20.625	20.5	3.889
El Salvador	SLV	38	35	34	43	43	42	46	44	40.625	42.5	4.406
Honduras	HND	29	17	20	24	30	33	34	43	28.75	29.5	8.345
Jamaica	JAM	.	.	.	.	.	.	.	.	.	.
Mexico	MEX	20	21	24	27	27	28	28	29.8	25.6	27	3.549
Nicaragua	NIC	55	32	29	32	33	32	32	31	34.5	32	8.367
Panama	PAN	57	49	42	43	40	37	35	35	42.25	41	7.573
Paraguay	PRY	18	15	13	14	12	10	11	13	13.25	13	2.493
Peru	PER	29.8	25.8	23.1	22.8	20.7	18.4	18	17	21.95	21.75	4.355
Uruguay	URY	61	57	45	53	40	38	40	37	46.375	42.5	9.349
Venezuela, RB	VEN	25	20	15	25	29	25	28	33	25	25	5.529

Measured in percentage of GDP.

Source: Economic Commission for Latin America and the Caribbean.

Table 3.

Inflation consumer price as annual percentage for 17 Latin American countries.

Country	Code	Year								Mean	Median	SD
Country	Code	2006	2007	2008	2009	2010	2011	2012	2013	Mean	Median	SD
Argentina	ARG	10.901	8.831	8.584	6.283	10.780	9.466	10.030	10.619	9.437	9.748	1.547
Bolivia	BOL	4.286	8.706	14.000	3.349	2.502	9.813	4.586	5.716	6.620	5.151	3.911
Brazil	BRA	4.184	3.637	5.663	4.886	5.038	6.636	5.402	6.202	5.206	5.220	0.994
Chile	CHL	3.392	4.408	8.716	0.072	1.410	3.340	3.007	1.792	3.267	3.173	2.589
Colombia	COL	4.296	5.544	6.997	4.203	2.278	3.412	3.177	2.023	3.991	3.807	1.664
Costa Rica	CRI	11.471	9.357	13.423	7.844	5.660	4.883	4.504	5.224	7.796	6.752	3.340
Ecuador	ECU	3.035	2.276	8.401	5.158	3.556	4.475	5.101	2.739	4.342	4.015	1.960
El Salvador	SLV	4.037	4.578	6.708	1.056	0.908	5.129	1.730	0.793	3.117	2.883	2.280
Honduras	HND	5.578	6.936	11.406	5.492	4.701	6.762	5.196	5.162	6.404	5.535	2.166
Jamaica	JAM	8.592	9.293	22.021	9.574	12.610	7.530	6.898	9.344	10.733	9.319	4.867
Mexico	MEX	3.629	3.967	5.125	5.297	4.157	3.407	4.112	3.806	4.188	4.039	0.679
Nicaragua	NIC	9.140	11.127	19.826	3.687	5.455	8.083	7.194	7.135	8.956	7.638	4.930
Panama	PAN	2.096	4.169	8.759	2.409	3.491	5.876	5.698	4.027	4.566	4.098	2.169
Paraguay	PRY	9.591	8.129	10.155	2.592	4.651	8.254	3.676	2.684	6.216	6.390	3.145
Peru	PER	2.002	1.780	5.786	2.936	1.528	3.371	3.654	2.817	2.984	2.876	1.363
Uruguay	URY	6.399	8.114	7.861	7.099	6.676	8.094	8.098	8.575	7.615	7.977	0.786
Venezuela, RB	VEN	13.663	18.699	31.441	27.081	28.187	26.090	21.069	40.639	25.859	26.586	8.277

Note. Measured in annual percentage.

Source: International Monetary Fund, International Financial Statistics.

This study revisits the findings from the above-mentioned studies with improved methods and up-to-date dataset.¹ In addition, this study breaks down crises (global crises, banking crises, currency crises, and debt crises), and poverty indicators (poverty gap measures and poverty headcount ratios with two indicators: US$1.90 a day and US$3.10 a day) into different types. The dataset used in this study is panel data that covers 17 Latin American countries between 1990 and 2013, which is by far the most-up-to date dataset in the literature. The methods used in this study are based on Hopkins (2006) and Nikoloski (2011) but are different from theirs because of the amended poverty indicators and the addition of global crisis as a dummy variable to investigate how the US-originated economic crisis in 2007 affected Latin American countries.

This study is of particular importance given that, although economic growth² is an important factor reducing a country’s poverty level, such link does not work in Latin American countries; economic growth of the several economies since the 2000s has little impact on reducing poverty and households with chronic poverty were unable to escape from poverty solely through economic growth (Vakis et al., 2016). Among many contributing factors, repeated economic crises in the region may be an important stumbling block. More specifically, economic crises cause aggregate demand to fall, and the relative price to change dramatically. This affects labor markets unfavorably, which decrease the demand in the labor markets and eventually decrease the real wage of workers (ILO 2017). Another important factor is the default in the financial system due to crises, which prevent the poor from accessing financial systems and they end up using the informal financial sector, which makes them even more vulnerable due to higher interest rate and reserve requirement (Gelos, 2009).

Data, methods, and models

This study covers 17 Latin American countries that have experienced at least one economic crisis between 1990 and 2013 (see Table 1 for the list of those countries). “Crisis” in this study is categorized into four different parts – banking, currency, debt, and global crisis – all of which have intrinsic links to poverty, as shown in Ocampo (2009) and Nikoloski (2011). Banking crisis is defined as the reduction in the value of the asset which often leads to bank closure, then government control over the financial system, which deteriorates the value of deposit and asset of the low-income family, which hinders the poor to access formal financial systems. Currency crisis is defined as the fluctuation in the exchange rate, foreign exchange reserve or short-term interest rates (Sharma, 1999), which not only influences the sharp decline in growth but also shifts the composition of the labor market, leading to the poor to work in low-quality employment (Nikoloski, 2011). Third, the debt crisis is defined as defaults on bank loans or bond spread above the critical level (Pescatori and Sy, 2007), which has a direct impact on government spending and lessens the investment in the social safety net. Lastly, a global crisis is defined as declined economic activity mainly due to external factors, such as trade shocks, which increases domestically the living cost of the low income group which is international financial linkage (Lane and Milesi-Ferretti, 2011; Ocampo, 2009).

The dependent variables in this study are four poverty indicators – two measurements (poverty gap³ and the component of the population considered to be poor) and two criteria (US$1.90 and US$3.10). The two different poverty lines are used in this study due to the sample heterogeneity; Brazil’s per capita GDP is much higher than that of Nicaragua, so it would be better to apply US$3.10 to Brazil and US$1.90 to Nigaragua to effectively measure the poverty level. The data source is the World Bank’s World Development Indicator and is primarily based on household survey data. Unlike the poverty gap that measures “depth” of poverty, headcount ratio simply measures the percentage of population under a certain poverty line (either US$1.90 or US$3.10 per day, with base year 2011).

The independent variables include four dummy variables for the economic crisis (banking, currency, debt, and global⁴), the initial macroeconomic condition, the percentage of foreign banks among total banks, bank deposit or national debt to GDP (Table 2), inflation and consumer price index (Table 3), trade as the percentage of GDP (Table 4), and general government final consumption expenditure. The regression equation is as follows

\begin{array}{l} P o v I n d e x_{i, t} = β_{0} + β_{1} D B a n k i n g_{i, t} + β_{2} D C u r r e n c y_{i, t} + β_{3} D D e b t_{i, t} + β_{4} D G l o b a l_{i, t} + \\ β_{5} l n G D P_{i, t} + β_{6} F b a n k_{i, t} + β_{7} D e p o s i t_{i, t} + β_{8} I n f l a t i o n_{i, t} + β_{9} T r a d e_{i, t} + \\ β_{10} G S p e n d i n g_{i, t} + α_{i} + ε_{i, t} \end{array}

where

$P o v I n d e x_{i, t}$ : the following four dependent variables with the base year 2011

PovGap1.90: income or consumption gap under the poverty line of US$1.90 a day

PovGap3.10: income or consumption gap under the poverty line of US$3.10 a day

HeadPov1.90: the percentage of the population that lives under the US$1.90 a day

HeadPov3.10: the percentage of the population that lives under the US$3.10 a day

Dummy independent variables are

$D B a n k i n g_{i, t}$ : banking crisis dummy variables

$D C u r r e n c y_{i, t}$ : currency crisis dummy variables

$D D e b t_{i, t}$ : debt crisis dummy variables

$D G l o b a l_{i, t}$ : global crisis dummy variables

The other independent variables are

$- l n G D P_{i, t}$ : log GDP per capita base on constant US dollars with baseline year of 2005

$F b a n k_{i, t}$ : percentage of foreign banks among total banks

$D e p o s i t_{i, t}$ : ratio of bank deposit (national debt) to GDP

$I n f l a t i o n_{i, t}$ : inflation value as consumer price in percentage, annual change in the cost of average consumer to obtain basket of good and service

$- T r a d e_{i, t}$ : trade as percentage of GDP (measure of openness of the country), sum of export and imports of goods and service divided by GDP

$G S p e n d i n g_{i, t}$ : government spending relates to government’s investment in public welfare, all expenditure of good and service in percent of the GDP

$α_{i}$ = country fixed effect

$ε_{i, t}$ = error term (i = 17 country, t = time periods 1990 to 2013)

Table 4.

Trade as percentage of GDP for 17 Latin American countries.

Country	Code	Year								Mean	Median	SD
Country	Code	2006	2007	2008	2009	2010	2011	2012	2013	Mean	Median	SD
Argentina	ARG	40.433	40.945	40.403	34.057	34.971	35.206	30.527	29.334	35.735	35.089	4.516
Bolivia	BOL	74.538	76.062	82.867	68.627	75.512	82.480	84.949	81.231	78.283	78.646	5.498
Brazil	BRA	26.042	25.293	27.258	22.106	22.517	23.702	24.765	25.562	24.656	25.029	1.769
Chile	CHL	73.099	76.407	80.790	66.337	69.064	72.206	68.272	64.973	71.393	70.635	5.331
Colombia	COL	38.174	36.330	38.052	34.280	33.701	38.669	38.296	37.698	36.900	37.875	1.931
Costa Rica	CRI	90.263	86.912	86.934	70.178	68.219	69.451	68.145	65.618	75.715	69.814	10.339
Ecuador	ECU	59.710	62.587	68.057	52.105	60.303	64.490	61.751	59.606	61.076	61.027	4.602
El Salvador	SLV	73.456	77.621	80.666	66.071	73.537	79.277	77.610	79.719	75.995	77.616	4.821
Honduras	HND	133.132	135.071	135.749	96.905	109.442	122.217	121.188	116.306	121.251	121.703	13.627
Jamaica	JAM	100.746	101.244	113.577	86.885	80.927	83.850	82.077	83.327	91.579	85.368	12.041
Mexico	MEX	56.093	56.795	57.777	55.968	60.760	63.470	65.767	63.765	60.049	59.269	3.905
Nicaragua	NIC	87.310	93.026	96.794	86.994	100.364	111.827	115.177	110.984	100.310	98.579	11.208
Panama	PAN	137.896	149.570	166.699	138.980	148.274	162.488	158.058	137.634	149.950	148.922	11.487
Paraguay	PRY	107.770	103.518	103.546	96.300	106.585	102.823	98.628	94.388	101.695	103.170	4.792
Peru	PER	51.785	55.688	58.434	48.112	51.673	55.988	52.620	49.777	53.010	52.202	3.451
Uruguay	URY	61.972	59.210	65.208	53.394	51.699	53.247	55.061	49.718	56.189	54.228	5.394
Venezuela, RB	VEN	58.666	56.199	51.829	38.521	46.137	49.638	50.404	54.278	50.709	51.116	6.303

Note. Measured in percentage of GDP.

Source: World Bank national accounts data, and OECD National Accounts.

Summary of descriptive statistics for each variable is provided in Table 5. Countries that did not go through any economic crisis at least once between 1990 and 2013 were excluded. In addition, due to the short data availability, data points before the year 1990 were not included. Since the poverty gap measure and the poverty headcount indicator are derived from income criteria, the multi-dimensional poverty perspective was not included. Other views, such as child mortality, nutrition intake, years of schooling, and child school attendance are excluded in income base poverty indicators. Moreover, housing-related factors such as availability of electricity, drinking water, sanitation, flooring, cooking fuel, and assets are also excluded.

Table 5.

Summary of descriptive statistics.

Variable	Observation	Mean	SD	Min	Max
Code(country)	408	9	4.905	1	17
Year	408	2001.5	6.931	1990	2013
PovGap1.90	408	4.547	3.776	0.06	20.77
PovGap3.10	408	8.822	5.99	0.35	33.84
HeadPov1.90	408	10.672	7.78	0.23	44.71
HeadPov3.10	408	21.044	12.542	1.56	63.24
Dbanking	408	0.12	0.325	0	1
Dcurrency	408	0.096	0.294	0	1
Ddebt	408	0.015	0.121	0	1
Dglobal	408	0.091	0.288	0	1
lnGDP	408	9.392	2.789	–2.567	16.524
Fbank	408	41.105	20.547	9	91
Deposit	408	32.165	17.889	4.443	104.967
Inflation	408	84.168	589.248	-1.167	7485.492
Trade	408	70.515	44.194	13.753	280.361
Gspending	408	12.604	3.54	2.976	43.479

As for the methods, this study adopts fixed effect (within transformation) panel analyses based on the Hausman test, stating that the fixed effect estimation is systematically different from random effect estimation, in which case, fixed effect provides consistent estimation results and overcome potential endogeneity problems.

Results

Table 6 shows the results for banking crisis. The coefficient for the banking crisis dummy is consistently positive for all the models (whether dependent variable is headcount ratio or poverty gap), implying that the banking crisis indeed exacerbates the poverty problem. Model 1-4, which uses headcount index US$3.10 a day, shows the largest coefficient of 4.482, stating that a banking crisis will increase the number of people living under this poverty line by 4.5%. In Model 1-1 where poverty gap of US$1.90 a day has been used as the dependent variable, the coefficient is 1.098, meaning that, when banking crisis occurs the gap between poverty line of US$1.90 a day and the lowest income group’s living cost becomes US$1.098 a day, which means the lowest income group in an economy would live with only US$0.802 a day. As for the other independent variables, logged GDP is significantly negative throughout all the models.

Table 6.

Fixed effects panel regression for banking crisis (Model 1-1 to 1-5).

Models	Model 1-1	Model 1-2	Model 1-3	Model 1-4	Model 1-5
Variables	PovGap1.90	PovGap3.10	HeadPov1.90	HeadPov3.10	HeadPov3.10
DGlobal
DBanking	1.098(0.40)***	2.162(0.58)***	2.668(0.77)***	4.482(1.17)***	3.794(1.09)***
DCurrency
DDebt
Trade	0.011(0.01)	0.005(0.01)	0.017(0.02)	0.005(0.02)	−0.011(0.02)
lnGDP	−0.857(0.12)***	−1.282(0.18)***	−1.701(0.24)***	−2.433(0.36)***	−0.303(0.43)
GSpending	0.165(0.05)***	0.110(0.07)	0.104(0.09)	−0.106(0.14)	0.383(0.13)
Inflation	0.000(0.00)	0.000(0.00)	0.000(0.00)	0.001(0.00)	0.000(0.00)
Deposit					−0.220(0.05)***
Fbank					−0.147(0.03)***
Number of observation	408	408	408	408	408
Number of groups	17	17	17	17	17

Standard errors in parenthesis are rounded up in third decimal and coefficients are rounded up in fourth decimal.

***indicates significant at 1%.

Table 7 explains currency crisis dummy. The poverty gap of US$1.90 a day did not provide any significant coefficient (the results have not been included in this table, though), and for this reason, this table focuses more on using the headcount ratio (four out of five models). The significantly positive signs shows distinctive worsening trends of poverty index when the currency crisis occurs, with larger coefficient than the case of banking crisis; the largest one is 8.437 in Model 2-5, which is large enough to take a number of population trapped in extremely poverty. As for the other variables, the coefficient for the logged GDP is significantly negative, while government spending, inflation, deposit, and foreign bank ratio do not show any significance with small values.

Table 7.

Fixed effects panel regression for currency crisis (Model 2-1 to 2-5).

Models	Model 2-1	Model 2-2	Model 2-3	Model 2-4	Model 2-5
Variables	PovGap3.10	HeadPov1.90	HeadPov1.90	HeadPov3.10	HeadPov3.10
DGlobal
DBanking
DCurrency	3.098(0.61)***	5.021(0.73)***	5.129(0.79)***	8.281(1.10)***	8.437(1.20)***
DDebt
Trade	−0.003(0.01)	−0.006(0.01)	0.005(0.01)	−0.032(0.02)	−0.015(0.02)
lnGDP	−1.293(0.17)***	−0.228(0.27)	−1.664(0.23)***	−0.158(0.41)	−2.378(0.34)***
GSpending	0.101(0.07)	0.170(0.08)**	0.080(0.09)	−0.006(0.13)	−0.144(0.14)
Inflation	0.000(0.00)	−0.000(0.00)	0.000(0.00)	−0.000(0.00)	0.001(0.00)
Deposit		−0.156(0.03)***		−0.244(0.46)***
Fbank		−0.087(0.02)***		−0.133(0.03)***
Number of observation	408	408	408	408	408
Number of groups	17	17	17	17	17

Standard errors in parenthesis are rounded up in third decimal and coefficients are rounded up in fourth decimal.

** indicates significant at 5%. *** indicates significant at 1%.

Tables 8 and 9 are about debt crisis and global crisis, respectively. Models 3-1 to 3-5 in Table 8 reveal that debt crisis does worsen the poverty. The effect is particularly large in Model 3-5 where US$3.10 has been used as a headcount poverty line – the coefficient for the dummy variable is as high as 8.952. The models in Table 9 provide interesting results for the global crisis dummy, which is negative. This unexpected result can be interpreted in several ways. For one reason, as the names tell, the global crisis was not originated domestically in Latin America but mixed with other economic factors, interpreting the link quite complicatedly; for example, before the external factors come into play in Latin American countries, they could have been able to cope with them with relevant policies, in which the actual poverty level in the region could have decreased. Other independent variables, however, have the same results as the previous ones.

Table 8.

Fixed effects panel regression for debt crisis (Model 3-1 to 3-5).

Models	Model 3-1	Model 3-2	Model 3-3	Model 3-4	Model 3-5
Variables	PovGap1.90	PovGap3.10	HeadPov1.90	HeadPov3.10	HeadPov3.10
DGlobal
DBanking
DCurrency
DDebt	3.167(0.98)***	5.908(1.39)***	5.661(1.95)***	12.143(2.75)***	8.952(2.98)***
Trade	0.004(0.01)	−0.008(0.11)	0.014(0.02)	−0.017(0.02)	0.001(0.02)
lnGDP	−0.280(0.15)*	−0.232(0.22)	−1.831(0.23)***	−0.233(0.43)	−2.654(0.36)***
GSpending	0.206(0.05)***	0.189(0.07)***	0.118(0.09)	0.067(0.13)	−0.081(0.14)
Inflation	−0.000(0.00)*	−0.000(0.00)	0.001(0.00)	−0.000(0.00)	0.013(0.00)**
Deposit	−0.082(0.02)***	−0.139(0.02)***		−0.243(0.05)***
Fbank	−0.025(0.01)**	−0.057(0.17)***		−0.162(0.03)***
Number of observation	408	408	408	408	408
Number of groups	17	17	17	17	17

Standard errors in parenthesis are rounded up in third decimal and coefficients are rounded up in fourth decimal.

indicates significant at 10%. ** indicates significant at 5%. *** indicates significant at 1%.

Table 9.

Fixed effects panel regression for global crisis (Model 4-1 to 4-5).

Models	Model 4-1	Model 4-2	Model 4-3	Model 4-4	Model 4-5
Variables	PovGap1.90	PovGap1.90	PovGap3.10	HeadPov1.90	HeadPov3.10
DGlobal	−1.837(0.41)***	−1.948(0.43)***	−3.001(0.59)***	−4.038(0.82)***	−6.610(1.25)***
DBanking
DCurrency
DDebt
Trade	0.008(0.01)	0.129(0.01)*	−0.002(0.01)	0.020(0.01)	0.015(0.02)
lnGDP	−0.245(0.15)	−0.799(0.12)***	−0.189(0.21)	−1.605(0.23)***	−2.082(0.33)***
GSpending	0.210(0.05)***	0.178(0.05)***	0.196(0.07)***	0.135(0.09)	−0.069(0.14)
Inflation	−0.000(0.00)*	−0.000(0.00)	−0.000(0.00)	0.001(0.00)	0.001(0.00)**
Deposit	−0.073(0.02)***		−0.123(0.02)***
Fbank	−0.023(0.01)*		−0.053(0.02)***
Number of observation	408	408	408	408	408
Number of groups	17	17	17	17	17

Standard errors in parenthesis are rounded up in third decimal and coefficients are rounded up in fourth decimal.

indicates significant at 10%. ** indicates significant at 5%. *** indicates significant at 1%.

Last, but not least, clustered standard error and correlation among dummy variables have been checked and to confirm there would be no multi-collinearity issue (see Table 10).

Table 10.

Correlation table for crisis dummy variables.

	Dglobal	Dbanking	Dcurrency	Ddebt
Dglobal	1	–	–	–
Dbanking	−0.1167	1	–	–
Dcurrency	−0.1027	0.3158	1	–
Ddebt	−0.0386	0.3307	0.3065	1

Conclusion

Many Latin American countries went through the so-called “lost decade,” with net external borrowing having skyrocketed and account deficits, as well as the trade deficits leading to debt accumulation and higher interest rates (Pastor, 1989). This may explain the significantly positive coefficients in debt crisis, larger than any other crisis dummies. Currency crisis also turned out to be an important factor while the banking crisis is not as crucial as expected. Global crisis brought unexpected results, which could be explained by several other factors tangled with it. Among other independent variables, the log GDP seems to be meaningful and significant throughout all the models.

After describing the impact of crises on poverty indicators, the next question should be how a country can recover from the poverty. This is particularly important in Latin America, as several IMF bailout package packages failed to be solutions and many countries had suffered from chronic and repeated crises. The adverse external conditions and the high poverty rate raised the cost of the IMF program implementation, which caused a high number of program cancellations and the countries resorted once again to IMF programs (Hutchison and Noy, 2003). Examining how to break this vicious cycle should be a very important and relevant follow-up topic, which is going to be reserved for further studies.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Ahmed

O’Donoghue

(2010) Global economic crisis and poverty in Pakistan. International Journal of Microsimulation 3(1): 127–129.

Breisinger

Diao

Collion

, et al (2011) Impacts of the triple global crisis on growth and poverty: The case of Yemen. Development Policy Review 29(2): 155–184.

Cline

(2002) Financial crises and poverty in emerging market economies. Center for Global Development Working Paper 8. Washington, DC: CGD. Available at: https://papers.ssrn.com/sol3/papers2.cfm?abstract_id=999990 (accessed 28 February 2017).

Dhanani

Islam

(2002) Poverty, vulnerability and social protection in a period of crisis: The case of Indonesia. World Development 30(7): 1211–1231.

Gelos

(2009) Banking spreads in Latin America. Economic Inquiry 47(4): 796–814.

Hopkins

(2006) Economic stability and health status: Evidence from East Asia before and after the 1990s economic crisis. Health Policy 75(3): 347–357.

Hutchison

Noy

(2003) Macroeconomic effects of IMF-sponsored programs in Latin America: Output costs, program recidivism and the vicious cycle of failed stabilizations. Journal of International Money and Finance 22(7): 991–1014.

International Labor Organization (2017) 2017 Labor overview of Latin America and the Caribbean. Geneva. Available at: http://www.ilo.org/wcmsp5/groups/public/—americas/—ro-lima/documents/publication/wcms_618120.pdf (accessed 27 July 2018).

Laeven

Valencia

(2008) Systemic banking crises: A new database. International Monetary Fund Working Papers 08/224. Washington, DC: Research Department, IMF. Available at: https://www.imf.org/external/pubs/ft/wp/2008/wp08224.pdf (accessed 28 February 2017).

10.

Laeven

Valencia

(2012) Systemic banking crises database: An update. International Monetary Fund Working Papers 12/163. Washington, DC: Research Department, IMF. Available at: http://nowandfutures.com/large/SystemicBankingCrisesDatabasewp12163(imf)pdf (accessed 28 February 2017).

11.

Lane

Milesi-Ferretti

(2011) The cross-country incidence of the global crisis. IMF Economic Review 59(1): 77–110.

12.

Lokshin

Ravallion

(2000) Welfare impacts of Russia’s 1998 financial crisis and the response of the public safety net. Economics of Transition 8(2): 269–295.

13.

Manning

(2000) Labour market adjustment to Indonesia’s economic crisis: Context, trends and implications. Bulletin of Indonesian Economic Studies 36(1): 105–136.

14.

Nikoloski

(2011) Impact of financial crises on poverty in the developing world: An empirical approach. Journal of Development Studies 47(11): 1757–1779.

15.

Ocampo

(2009) Latin America and the global financial crisis. Cambridge Journal of Economics 33(4): 703–724.

16.

Pastor

(1989) Latin America, the debt crisis, and the International Monetary Fund. Latin American Perspectives 16(1): 79–110.

17.

Pescatori

(2007) Are debt crises adequately defined? IMF Staff Papers 54(2): 306–337.

18.

Ravallion

Shaohua

Prem

(2009) Dollar a day revisited. The World Bank Economic Review 23(2): 163–184.

19.

Sharma

(1999) The challenge of predicting economic crises. Finance and Development 36(2): 40.

20.

Suryahadi

Sumarto

(2003) Poverty and vulnerability in Indonesia before and after the economic crisis. Asian Economic Journal 17(1): 45–64.

21.

Vakis

Rigolini

Lucchetti

(2016) Left behind: Chronic poverty in Latin America and the Caribbean. Washington, DC: International Bank for Reconstruction and Development, WB. Available at: https://openknowledge.worldbank.org/bitstream/handle/10986/21552/9781464806605.pdf?sequence=10 (accessed 27 July 2018).