ArticlePDF Available

The effect of reducing the pre-harvest burning of sugar cane on respiratory health in Brazil

Authors:

Abstract and Figures

This paper analyzes the impact of the increase in the raw sugar-cane har-vest area on the population respiratory health in Brazil. We collect data for the São Paulo state municipalities for two different periods: 2000, before state law defining gradual elimination of pre-burning sugar-cane area and 2007. We used panel models for inpatient visits due to respiratory diseases, outpatient visits for inhalation procedure and length of stay for inpatient visits due to respiratory diseases, controlling for the endogeneity between health and pol-lution. Results show that the increase of raw sugar-cane harvest area reduces the number of inpatient visits and do not impact the number of inhalation and length of stay.
Content may be subject to copyright.
Environment and Development
Economics
http://journals.cambridge.org/EDE
Additional services for Environment and
Development Economics:
Email alerts: Click here
Subscriptions: Click here
Commercial reprints: Click here
Terms of use : Click here
The effect of reducing the pre-harvest burning of
sugar cane on respiratory health in Brazil
Alexandre C. Nicolella and Walter Belluzzo
Environment and Development Economics / Volume 20 / Issue 01 / February 2015, pp 127 -
140
DOI: 10.1017/S1355770X14000096, Published online: 01 April 2014
Link to this article: http://journals.cambridge.org/abstract_S1355770X14000096
How to cite this article:
Alexandre C. Nicolella and Walter Belluzzo (2015). The effect of reducing the pre-
harvest burning of sugar cane on respiratory health in Brazil. Environment and
Development Economics, 20, pp 127-140 doi:10.1017/S1355770X14000096
Request Permissions : Click here
Downloaded from http://journals.cambridge.org/EDE, IP address: 189.63.233.34 on 05 Jan 2015
Environment and Development Economics 20: 127–140 © Cambridge University Press 2014
doi:10.1017/S1355770X14000096
The effect of reducing the pre-harvest burning of
sugar cane on respiratory health in Brazil
ALEXANDRE C. NICOLELLA
Department of Economics, University of S˜ao Paulo at Ribeir˜ao Preto, Av.
Bandeirantes 3900, FEARP, Ribeir˜ao Preto, S˜ao Paulo 14040-900, Brazil.
Email: anicolella@fearp.usp.br
WALTER BELLUZZO
Department of Economics, University of S˜ao Paulo at Ribeir˜ao Preto, Brazil.
Email: belluzzo@usp.br
Submitted 18 October 2012; revised 30 August 2013; accepted 31 January 2014; first published
online 1 April 2014
ABSTRACT. This paper analyzes the effect of reducing pre-harvest burning of sugar cane
on the population’s respiratory health in Brazil. We collected data for the municipalities
in the state of S˜
ao Paulo for two different periods: 2000, before the state law requiring the
gradual elimination of sugar cane area utilizing pre-burning, and 2007. We used panel
models for inpatient visits due to respiratory diseases, outpatient visits for inhalation
procedures and length of stay for inpatient visits due to respiratory diseases, controlling
for the endogeneity between health and pollution. The results show that increasing the
area of raw sugar cane harvested reduces the number of inpatient visits and does not
affect the number of inhalation procedures or length of stay.
1. Introduction
Concerns regarding global warming and the prospect of a future scarcity of
fossil fuels brought the use of biofuel into the energy policy agenda. As a
result, biofuel is gaining importance in the world energy matrix, and there
are projections that it will account for 5 per cent of liquid fuels by 2025 (EIA,
2006). In this context, ethanol arises as one of the best alternatives among
biofuels.
The author thanks the German Federal Ministry for Economic Cooperation and
Development – BMZ, German Development Institute – DIE, In Went and Keynes
College at the University of Kent.
128 Alexandre C. Nicolella and Walter Belluzzo
The use of sugar cane ethanol as a substitute for gasoline is widespread
in Brazil.1The success of the Brazilian program is due to several factors.
On the supply side, there were significant cost reductions over the last
several decades due to heavy investments in technology that increased
sugar cane productivity from 34 t.ha1in 1960 to 69 t.ha1in 2006 (IBGE,
2010a). On the demand side, there were three important factors: the imple-
mentation of flex-fuel technology, which allows automobiles to run on any
proportion of ethanol and gasoline2, the mandatory blend of ethanol and
gasoline,3and, perhaps more importantly, the increase in oil prices.
The effect of these shifts on supply and demand in the recent decades
caused ethanol production and the sugar cane area to soar. In 2009, Brazil
was responsible for about one–third of the world production of ethanol,
while the area grew from 4.9 million ha in 2000 to 8.2 million ha in 2008,
a growth rate of 6.7 per cent per year. Such a rapid expansion certainly
has had a social and environmental impact, which has been addressed
in the literature. There are studies on soil degradation (Giampietro et al.,
1997;Moreira and Goldemberg,1999;Oliveira et al.,2005), water pollution
(Moreira and Goldemberg,1999;Gunkel et al.,2007) and greenhouse gases
emissions (Moreira and Goldemberg,1999;Moreira,2000;Oliveira et al.,
2005;Crutzen et al.,2008;B¨
orjesson,2009), and air pollution (Allen et al.,
2004;Lara et al.,2005;Andrade et al.,2010).
The connection between sugar cane ethanol production and air pollu-
tion is the harvesting of sugar cane. Sugar cane is a semi-perennial culture
and has two different harvest systems: the mechanical harvest of raw sugar
cane or the manual harvest of previously burned sugar cane. Because large
areas are burned at once, this sort of harvesting has a significant impact on
air pollution. In fact, there are studies presenting evidence that pre-harvest
burning is responsible for the increase of fine particulate matter, coarse
particulate matter and black carbon concentrations, especially in the time
period the burning occurs (Lara et al.,2005), which increases concentrations
of substances such as nitrites, sulfites, carbon oxide and other substances
(Allen et al.,2004). As indicated in the literature, short- and long-term expo-
sure to these pollutants can negatively affect human health capital (Sicard
et al.,2010), especially for the young and elderly (Braga et al.,1999;Farhat
et al.,2005;Gonc¸alves et al.,2005;Roseiro and Angela,2006).
Although the pollution from sugar cane pre-harvest burning may be as
harmful as traffic and industrial pollution (Mazzoli-Rocha et al.,2008), there
are few studies addressing the effect of the pre-harvest burning of sugar
cane on health. Our literature review revealed only a handful of studies
from recent years: Uriarte et al. (2009), Ribeiro (2008), Arbex et al. (2007),
1In 2009, Brazilian consumption of ethanol reached 11.3 million tonnes of oil equiv-
alent (toe), compared to 14.7 million toe of gasoline. Additionally, sugar cane
accounts for 18.1 pre cent of the Brazilian energy matrix (EPE,2010).
2The percentage of flex-fuel small-sized cars sold in Brazil in 2005 was 39 per cent
rising to 87 per cent in 2009 (ANFAVEA,2010).
3Brazil produces an anhydrous ethanol (99.6 GL) as an octane enhancer in gaso-
line with blending rates that range from 20 per cent to 26 per cent, and hydrated
ethanol (95.5 GL) for neat-ethanol engines and flex fuel engines
Environment and Development Economics 129
Canc¸ado et al. (2006), and Arbex et al. (2000). Most of these studies are
restricted to the effect in specific municipalities and only Uriarte et al. (2009)
considers a larger geographic area. Moreover, these studies do not present
a well-defined identification strategy.
The identification problem of accessing the effect of the pre-harvest burn-
ing of sugar cane on respiratory health is the bias arising from unobserved
heterogeneity and endogeneity of the total sugar cane area. Because total
sugar cane area is likely to be endogenous in the equation for respiratory
diseases, and the parameter of interest cannot be estimated consistently, it
is therefore not identified.
The main contribution of this article is to present an estimation strat-
egy that addresses unobserved heterogeneity and endogeneity, considering
data from a larger geographic area and a long time span. In particular, we
address unobserved heterogeneity using panel data methods and avoid
endogeneity by using the change in the area harvested mechanically, with-
out fire, instead of the total area harvested. We argue that exogeneity of
the area harvested mechanically follows from the enactment of a state law
establishing a schedule for the gradual elimination of pre-harvest burning.
Additionally, this paper contrasts with the existing literature by con-
trolling for other sources of pollution, such as the automobile fleet and
industrial production, socioeconomic variables, and the extent of the pri-
vate sector health supply system. The introduction of additional control
variables and the consideration of inpatient and outpatient visits separately
provides a clearer picture of the effect of sugar cane harvesting on health.
To achieve these objectives, we collected data for the municipalities in
the state of S˜
ao Paulo, Brazil, for two different periods, 2000 and 2007,
and estimate a fixed effects model considering three different proxies for
health: hospital admissions due to respiratory conditions; the number of
inhalation procedures; and the average length of hospitalization due to a
respiratory condition.
2. Methodology
2.1. Estimation strategy
The main estimation problem to access the effect of sugar cane pre-harvest
burning on respiratory health is the bias arising from the unobserved het-
erogeneity and endogeneity of the total sugar cane area. Total sugar cane
area is led by economic growth, among other factors, which is likely to be
related to other sources of air pollution and/or the quality of the health
system. Thus, the change in the total sugar cane area is endogenous in the
equation for respiratory diseases, and the coefficient of interest cannot be
estimated consistently.
Our strategy to tackle the endogeneity problem is to estimate a fixed
effects panel data model using the change in the area harvested mechani-
cally instead of the total area. The changes in the area harvested mechani-
cally are induced by the enactment of S˜
ao Paulo State Law 11,241 in 2002.
This state law requires farms with less than a 12 pre cent slope to com-
pletely eliminate burning by 2021, starting with a 20 per cent reduction
in the burned area immediately. Farms with a greater than 12 per cent
130 Alexandre C. Nicolella and Walter Belluzzo
slope must completely eliminate buring by 2031, starting with a 10 per cent
reduction in 2002.4
Producers started adapting rapidly to State Law 11,241, and in 2007
approximately 47 per cent of the sugar cane area in the state of S˜
ao Paulo
was already harvested without burning (SMA,2010). Apparently, such
rapid adaptation facilitated accelerating the deadlines for the abolishment
of pre-harvest burning to 2014 and 2017, in a protocol signed in 2007 by the
S˜
ao Paulo State Secretary of the Environment and the Brazilian Sugar Cane
Industry Association (Uni˜ao da Ind ´ustria da Cana-de-A¸ucar – UNICA).
We argue that the change in the area harvested mechanically induced
by State Law 11,241 is exogenous in the equation for respiratory health
because it is not related to factors such as economic growth. The reasoning
is that the changes in area across municipalities depend basically on geo-
graphic characteristics, such as terrain slope, which are independent of the
change in variables measuring respiratory diseases.
Because all municipalities are subject to the law, but it imposes differ-
ent requirements for areas with slopes greater than 12 per cent, assuming
exogeneity of the changes in the area harvested mechanically implies
assuming that slopes are not related to respiratory health through unob-
served factors. Thus we assume that respiratory health is conditionally
not correlated to the area harvested mechanically or the terrain slope.
Any unobserved factors that may be correlated are assumed to be time
invariant, such that the coefficient of interest can be estimated consistently
through a fixed effects model.
2.2. Data
As discussed before, our estimation relies on the variation in the area har-
vested of sugar cane without burning. This variation was induced by the
S˜
ao Paulo state law, enacted in 2002. For this reason, we limit our analy-
sis to the state of Sao Paulo. It is worth noting that the state is the main
producer of sugar cane, corresponding to 60 per cent of the total Brazil-
ian production and 55 per cent of the total sugar cane area in 2008 (IBGE,
2010b).
We collected data from the 643 municipalities for 2000, before the state
law, and 2007, five years later. We assume that in 2000 only pre-harvest
burning was used.
In order to provide a more nuanced analysis of the effect of pre-harvest
burning on health, three proxy variables for health were used. The first
proxy is the number of inpatient visits or hospital admissions due to
a respiratory condition. The second dependent variable is the number
of outpatient visits due to the necessity of inhalation procedures. The
third variable is the average length of hospitalization due to a respiratory
condition.
Data on these proxy variables were collected from the database of the
Brazilian Ministry of Health (DATASUS), according to the International
4According to (Aguiar et al.,2009), about 25 per cent of the total area harvested has
slopes greater than 12 per cent.
Environment and Development Economics 131
Classification of Diseases 10th codes j00 to j99 (DATASUS,2010). Because
the observational units are municipalities, visits were measured per 1,000
inhabitants, per year. Additionally, because individuals living in smaller
cities might seek treatment in neighboring cities, we accounted for the
patient’s city of residence to build the variables.
A shortcoming of using the Ministry of Health data is that only the free
public health system visits are accounted for. This is important, because
the private health system is relatively large in the State of S˜
ao Paulo,5
and the distribution between public and private systems varies consid-
erably among municipalities. If patients using the private system were
independent and identically distributed (iid), the share would be constant,
facilitating the correction required to make inferences about the whole
population. As discussed later, handling this complication requires an
additional hypothesis to define the proper correction.
Considering that the characterization of air pollution is usually linked to
the concentration of pollutants in the atmosphere, we use three indicators
to account for the volume of pollutants per m3of air, each controlling a
different source of pollution: (i) area in thousands of hectares of raw sugar
cane that was harvested without pre-burning (IBGE,2010b;SMA,2010);
(ii) total fleet in 2002 as a proxy 2000 fleet, in thousands (SEADE,2010);
and (iii) the 2001 total consumption of industrial energy, as a proxy for
2000’s consumption, in 1,000 MWh (SEADE,2010). Implicit is the hypoth-
esis that energy consumption is related to production, which is related to
the amount of pollution generated.
Because the number of visits and length of hospitalization are likely to
be correlated with environmental factors facilitating the spread of viruses
and with the age of the population, we control for the municipalities’ pop-
ulation densities and the percentage of the population above 60 years old
to control for age (SEADE,2010).
For the investment in health, two variables were used: the number of
health professionals and nurses per 1,000 inhabitants registered in the
regional council for each municipality (SEADE,2010) and the percent-
age of the population covered by private health insurance, coded in three
categories: all ages, 14 years old or younger and 60 years old or older
(DATASUS,2010).
Due to the fact that wages and education have a high correlation, we use
the average of real salaries for formal workers to account for both factors
that can affect the capacity to invest in health and the efficiency of such
investment (SEADE,2010).
Although the weather is an important variable for explaining respiratory
conditions, the yearly average temperature, precipitation and humidity do
not vary substantially for each city from 2000 to 2007 (CIIAGRO,2010).
For this reason, we assume that the annual average climate variables are
constant variables over time, so that heterogeneity is captured by the panel
data model.
5In 2007, approximately 38 per cent of the state’s population was covered by
private insurance.
132 Alexandre C. Nicolella and Walter Belluzzo
2.3. Econometric specification
To circumvent the problem of the unobserved heterogeneity and endogene-
ity of the total sugar cane area presented above, we proposed a panel data
model to estimate the effect of increasing the area of raw sugar cane harvest
as follows:
hit =β1c
it +β2o
it +β3Sit +β4Mit +β5Wit +εit
εit =uit +θi,and uit N(0,1)(1)
where irepresents the 643 municipalities and tthe two periods, 2000 and
2007. The variable h
it is the number of inpatient visits per 1,000 inhabitants,
the number of inhalation procedures per 1,000 inhabitants, and the average
length of stay. c
it is the harvest area of raw sugar-cane, o
it is the total
fleet and total industrial energy consumption, Sit is the population density
and percentage of the population above 60, Mit is the number of health
professionals and nurses per thousand inhabitants and the percentage of
the population covered by health insurance, and Wit is the average salary.
The error term uit is homoscedastic and non-autocorrelated and θiis the
time invariant individual and municipality specific effects, such as weather
and individual behavior in the presence of pollution.6
3. Results
Descriptive statistics for the variables used in the models are presented in
table 1, and the correlations among independent variables are presented in
table 2. We note that these correlations are relatively small, even though
there are some correlation coefficients close to |0.60|. Nonetheless, stability
of the estimated coefficients and the results obtained for the t-andF-tests
suggest that multicollinearity is not a concern.
To analyze model sensitivity, we successively added blocks of inde-
pendent variables and evaluated the changes in magnitude, sign and
significance of the parameter estimates. Table 3presents the fixed effect
regressions with cluster robust standard errors for inpatient visits due to
respiratory diseases per 1,000 inhabitants.
The first equation uses just the environmental factors, the second adds
the pollution factors, the third adds health goods, and the fourth is the
complete equation with salary. We observe that the magnitudes, signs and
significance of the parameter estimates are reasonably stable. For instance,
the effect of area of raw sugar-cane harvested on inpatient visits was 0.172
without controlling for health goods, education or salary and a decrease
in module to 0.129 for the complete model. So, by not controlling for
6An alternative econometric specification for the error term is to consider the
spatial dependence among neighboring municipalities. In this particular case,
because there is no spatially lagged dependent variable in the model, ignoring
spatial dependence does not produce any bias, affecting only the variance of
the estimated coefficients. We chose to use clustered robust standard errors as
a variance correction.
Environment and Development Economics 133
Table 1. Summary statistics for the variables used in the models
Variable Mean Std. Dev. N
Inpatient visits/1,000 inhabitants 11.143 7.472 1288
Inpatient visits with private insurance/1,000 inhabitants 12.939 9.374 1288
Inpatient visits <15 years/1,000 inhabitants 18.347 12.669 1288
Inpatient visits >60 years/1,000 inhabitants 30.105 20.064 1288
Inhalation procedure/1,000 inhabitants 442.449 382.447 1288
Average length of stay (days) 5.031 1.465 1288
Population density 284.069 1141.331 1288
Population <15 years old 0.249 0.038 1288
Population >60 years old 0.112 0.028 1288
Harvest w/burning (1,000 ha) 1.411 3.617 1288
Total fleet (1,000 unit) 22.118 194.219 1288
Industry energy (1,000 MWh) 71.850 320.318 1286
Health professionals/1,000 inhabitants 4.799 2.782 1288
Nurses/1,000 inhabitants 4.18 2.373 1288
Private insurance/population 0.151 0.153 1288
Private insurance/population <15 years 0.146 0.167 1288
Private insurance/population >60 years 0.155 0.141 1288
Salary (2007 R$)a1052.553 320.991 1288
Note:aThis series is in real price of 2007 deflated by INPC (National Price
Consumer Index) for income (IPEADATA, 2010).
Table 2. Linear correlation among variables used in the models
Pop. Pop. Raw Total Ind. Health Private
Variables Density >60 sugar-cane fleet energy profes. benef. Salary
Population density 1.000
Population >60 0.265 1.000
Harvest w/burning 0.068 0.075 1.000
Total fleet 0.303 0.045 0.000 1.000
Industrial energy 0.310 0.183 0.007 0.582 1.000
Health professionals 0.191 0.149 0.188 0.188 0.200 1.000
Private insurance 0.314 0.214 0.160 0.197 0.415 0.332 1.000
Salary (2007 R$) 0.367 0.355 0.113 0.226 0.517 0.295 0.587 1.000
those factors, one might overestimate the effect of increasing the area of
raw sugar cane harvest.
Considering the possibility of influential observations, we identified the
city of S˜
ao Paulo as the only obvious candidate. With a population of
around 11 million and being the largest city in South America, S˜
ao Paulo
holds a considerable share of the Brazilian industry and of Brazil’s vehi-
cle fleet, which is certainly unlike any other Brazilian city. For this reason
we also estimated the complete model, dropping the city of S˜
ao Paulo, as
shown in the last column of table 3. Comparing the parameter estimates
134 Alexandre C. Nicolella and Walter Belluzzo
Table 3. Sensitivity analysis of the panel data model for inpatient visits due to
respiratory condition
Dependent variable – inpatient visits per 1,000
inhabitants due to respiratory diseases
Independent StSt,St,,MtSt,,Mt,Et,WtSt,,Mt,Et,Wt
variables FE model FE model FE model FE model FE model–SP
Population 0.00121 0.000615 0.00145 0.00101 0.00118
density (0.00122) (0.00124) (0.00120) (0.00123) (0.00118)
Population >60 147.6∗∗∗ 125.4∗∗∗ 75.54∗∗ 66.84∗∗ 66.36∗∗
(21.37) (23.69) (31.93) (31.72) (31.91)
Harvest 0.172∗∗∗ 0.140∗∗∗ 0.129∗∗∗ 0.127∗∗∗
w/burning (0.0379) (0.0355) (0.0355) (0.0355)
Total fleet 0.00269∗∗∗ 0.00342∗∗∗ 0.00267∗∗∗ 0.00246
(0.000949) (0.000893) (0.000955) (0.0154)
Industrial 0.00128∗∗∗ 0.00159∗∗∗ 0.00161∗∗∗ 0.00168∗∗∗
energy (0.000421) (0.000457) (0.000423) (0.000428)
Health 0.492∗∗ 0.480∗∗ 0.476∗∗
professionals (0.194) (0.191) (0.192)
Private 3.860∗∗ 3.4863.520
insurance (1.928) (1.904) (1.918)
Salary (2007 R$) 0.00297∗∗ 0.00301∗∗
(0.00134) (0.00136)
Constant 27.27∗∗∗ 25.07∗∗∗ 22.12∗∗∗ 24.29∗∗∗ 24.32∗∗∗
(2.300) (2.514) (2.864) (2.933) (2.933)
Observations 1,288 1,286 1,286 1,286 1,284
R-squared 0.113 0.130 0.145 0.152 0.152
Number of
municipalities
644 643 643 643 642
Notes: Clustered robust standard errors in parentheses.
∗∗∗p<0.01; ∗∗p<0.05; p<0.1.
with the previous estimates we can see that, in general, they agree in both
sign and significance and magnitude, except for ‘total fleet’.
3.1. Inpatient visits
We specified four different models to account for the effect of the area of
raw sugar cane harvest on inpatient visits due to respiratory diseases. The
estimated models are presented in table 4. According to Hausman tests,
presented at the bottom of table 4, we reject at 1 per cent the null hypothesis
that the differences are in the coefficients of both models, suggesting that a
fixed effects model is more appropriate.
The first model is for the number of inpatient visits, controlling for
the percentage of individuals covered by private health insurance. In the
second model, we assume that all inpatient visits correspond to uninsured
individuals, covered only by the free public health system. We assume
further that visits are equally distributed for both public and private sys-
tems. Therefore, the total number of visits were approximated by dividing
Environment and Development Economics 135
Table 4. Estimation results for distinct measures of inpatient visits due to
respiratory conditions
Dependent variable – inpatient visits
due to respiratory diseases
Inpatient with Inpatient Inpatient
Inpatient private sector <15 visits >60
Variable Fixed effect model
Population density 0.00101 0.00133 0.00282 0.00229
(0.00123) (0.00215) (0.00214) (0.00360)
Population >60 66.84∗∗ 66.94∗∗
(31.72) (33.88)
Harvest w/burning 0.129∗∗∗ 0.148∗∗ 0.192∗∗ 0.378∗∗∗
(0.0355) (0.0678) (0.0772) (0.103)
Total fleet 0.00267∗∗∗ 0.00203 0.00271 0.00609∗∗∗
(0.000955) (0.00349) (0.00283) (0.00233)
Industrial energy 0.00161∗∗∗ 0.0168∗∗∗ 0.00203∗∗ 0.00336∗∗∗
(0.000423) (0.00203) (0.000864) (0.000860)
Health professionals 0.480∗∗ 0.628∗∗∗ 0.871∗∗∗ 1.737∗∗∗
(0.191) (0.221) (0.241) (0.429)
Private insurance 3.486
(1.904)
Private insurance <15 2.275
(2.560)
Private insurance >60 4.507
(8.136)
Salary (2007 R$) 0.00297∗∗ 0.002950.00294 0.00863∗∗
(0.00134) (0.00159) (0.00203) (0.00419)
Constant 24.29∗∗∗ 25.87∗∗∗ 25.24∗∗∗ 49.06∗∗∗
(2.933) (3.236) (2.341) (4.641)
F19.24∗∗∗ 20.14∗∗∗ 6.27∗∗∗ 13.25∗∗∗
Observations 1286 1286 1286 1286
Hausman test 60.40∗∗∗ 53.15∗∗∗ 32.20∗∗∗ 34.69∗∗∗
Notes: Clustered robust standard errors in parentheses.
∗∗∗p<0.01; ∗∗p<0.05; p<0.1.
the observed number of visits in the public system by the proportion of the
population not covered by private insurance.
The third model considers the number of inpatient visits by individuals
younger than 15 years old, and the last model considers the number of vis-
its for the population older than 60 years old. Note that in these last two
models, there is no correction to take into account visits in the private sys-
tem, but they do include controls for the private coverage share, computed
within each age bracket.
The variable of interest, the raw sugar cane harvest area (c), has a
negative and significant effect in all estimated models. The increase of
1,000 hectares harvested without burning decreases the number of inpa-
tient visits by 0.129 per 1,000 inhabitants. Correcting the number of visits
136 Alexandre C. Nicolella and Walter Belluzzo
for private coverage, the estimated effect increases, in absolute value, to
0.148. The effect of the area of raw sugar cane harvest is more important for
the population younger than 15 years old and older than 60 years old. The
effect of an increase of 1,000 hectares of sugar cane harvested without burn-
ing decrease the inpatient visits by 0.192 and 0.378, respectively. Finally,
it is worth noting that those coefficients are lower than the ones shown
in table 3without controlling for health goods, indicating that omitting
control variables may result in overestimating effects.
These results relating the pre-harvest burning to health are in line with
the literature. Uriarte et al. (2009), Arbex et al. (2007), Canc¸ado et al. (2006)
and Arbex et al. (2000), for example, present evidence of a positive relation-
ship between pre-harvest burning and respiratory diseases. The literature
on the relationship between urban air pollution and health is much more
extensive, with a considerable number of studies presenting evidence in
line with our results, such as Farhat et al. (2005); Gonc¸alves et al. (2005);
Roseiro and Angela (2006); Sicard et al. (2010)andBraga et al. (1999).
Even though the controlling variables are included in the model mainly
to isolate the effect of the variable of interest, the corresponding parameter
estimates may shed some light on the internal validity of the models. An
inspection of table 4reveals that most parameter estimates for the control-
ling variables have the correct sign in most cases. An obvious exception is
that the percentage of the population over 60 years old has a significant
negative effect in the models regardless of correcting for private coverage.
In other words, an increase in the population over 60 decreases the number
of inpatient visits per 1,000 inhabitants.
For the variable controlling for other sources of pollution (o),we
observe that the coefficient for the total fleet is positive and significant
for the model with the sample restricted to those older than 60 years old.
Industrial energy consumption has a positive and significant effect in all
the estimated models, with a slightly greater effect for the sample restricted
to older individuals.
The number of health professionals has a negative and significant effect
on the number of inpatient visits. This may be expected because better
supply conditions facilitates access to health care, which in turn leads to
less acute problems requiring hospitalization. On the other hand, more
health goods imply better health conditions. This is especially important
for individuals younger than 15 and over 60 years old. Thus, according to
the inpatient visits model, an increase of one health professional per 1,000
inhabitants decreases the inpatient visits by 0.48 for the whole population,
0.628 for inpatient care in the private health sector model, 0.871 for the
population younger than 15 years old, and 1.737 for the population over 60
years of age.
The percentage of the population in the municipalities covered by pri-
vate health insurance has a negative and significant effect (at the 10 per cent
level) in the model considering the whole population, and is not signif-
icant for the models with their samples restricted to younger and older
individuals.
Finally, the average salary (W) in a municipality has a negative effect on
inpatient visits. The argument here is that better salaries are correlated with
Environment and Development Economics 137
Table 5. Estimation results for outpatient visits and length of stay due to respiratory
conditions
Inhalation Inhalation Length of stay
Variable FE Model RE Model FE Model
Population density 0.0315 0.00584 0.000421
(0.0523) (0.00956) (0.000385)
Population <15 2,043∗∗∗ 1,633∗∗∗
(627.5) (298.7)
Population >60 10.59
(5.470)
Harvest w/burning 0.701 1.770 0.0106
(2.707) (2.327) (0.00996)
Total fleet 0.0725 0.0167 0.000690
(0.0882) (0.0418) (0.000657)
Industrial energy 0.122∗∗ 0.0434 0.000350
(0.0483) (0.0411) (0.000264)
Nurses 6.053 11.17∗∗∗
(9.986) (4.106)
Health professionals 0.0458
(0.0402)
Private insurance 12.10 220.6∗∗ 0.456
(153.0) (109.0) (0.518)
Salary (2007 R$) 0.0316 0.07940.000397
(0.0781) (0.0471) (0.000274)
Constant 61.17 136.6 6.166∗∗∗
(222.4) (86.04) (0.554)
F7.78∗∗∗ 1535.74∗∗∗ 1.55
Observations 1286 1286 1286
Hausman test 9.52 9.52 12.19∗∗
Notes: Clustered robust standard errors in parentheses.
∗∗∗p<0.01; ∗∗p<0.05; p<0.1.
better education, implying a greater and more efficient capacity to invest in
health and so fewer inpatient visits. The effect of salary was higher for the
population over 60 years old and it was not significant for the population
under 15 years of age.
3.2. Outpatient visits and length of stay
Tabl e 5shows the estimation results for the other two proxies for health,
namely outpatient visits and length of hospitalization due to a respiratory
condition. The model for outpatient visits includes the number of inhala-
tion procedures per 1,000 inhabitants as the dependent variable, and the
corresponding estimates are shown in the first column of table 5.The
second covers inpatient length of stay due to respiratory diseases.
Unlike the model for inpatient visits, the null hypothesis of the Haus-
man test is not rejected for the outpatient visits model. Rejection of the null
138 Alexandre C. Nicolella and Walter Belluzzo
hypothesis suggests that the random effects model may be more appropri-
ate in this case. Nonetheless, because the random effects model is difficult
to justify in this setting, we present both fixed and random effects models.
For the length of stay model, the null hypothesis is rejected, and thus only
the fixed effects model is presented. See the test results at the bottom of
table 5.
Starting with the variable of interest, the area of raw sugar cane harvest,
it is clear from the results presented that there is no statistically significant
effect on either outpatient visits or length of hospitalization. The contrast
of this finding with the results obtained earlier for inpatient visits sug-
gests that pre-harvest burning imposes a harsher toll on acute respiratory
conditions, which require brief hospitalization and reduced variance in
the length of stay. It also indicates that focusing on outpatient visits for
inhalation may underestimate the health effect of air pollution.
Also contrasting with the model for inpatient visits, few of the control
variables have statistically significant coefficients. The model for length of
stay did not reveal any interesting health effect associated with the control
variables, nor did the model for the number of outpatient visits. In this case,
the proportion of the population younger than 15 years old seems to hold
the main effect, with a positive and significant effect on outpatient visits, in
both the random effects and fixed effects models.
4. Conclusion
This paper presented an analysis of the respiratory health effect of the
air pollution caused by the pre-harvest burning of sugar-cane. As sugar
cane has increased in importance in the energy agenda in Brazil and in the
world, it is important to shed some light on the impact of its production on
the environment and the health of the population. This article contributes
to this body of literature by presenting evidence that there may be a sig-
nificant health effect, using a novel identification strategy and a unique
data set.
The proposed identification strategy relies on the enactment of a state
law reducing the pre-harvest-burned area over time. After collecting data
before and after the law took effect, we estimate a series of panel data mod-
els that control for endogeneity and make it possible to evaluate the health
effect of reducing the area burned.
The results obtained suggest that reducing the area where sugar cane is
harvested after burning reduced the number of inpatient visits in the state
of S˜
ao Paulo. Interestingly, we found that the effect of pre-harvest burning
is relatively large, as compared to the estimated effects of the total vehicle
fleet and industrial pollution, included as control variables in the models.
On the other hand, the models estimated for the number of outpatient
visits for inhalation procedures and for the length of stay did not reveal
any significant relationship between pre-harvest burning and respiratory
conditions.
These results suggest that the effect may be restricted to acute respiratory
conditions, which require brief hospitalization and reduced variance in the
Environment and Development Economics 139
length of stay. Moreover, based on these results, we may argue that the state
law contributed to the improvement of the population’s respiratory health
in the state of S˜
ao Paulo.
References
Aguiar, D.A., B.F.T. Rudorff, M. Adami, and Y.E. Shimabukuro (2009), ‘Imagens
de sensoriamento remoto no monitoramento da colheita da cana-de-ac¸´
ucar’,
Engenharia Agr´ıcola 29(3): 440–451.
Allen, A., A. Cardoso, and G. da Rocha (2004), ‘Influence of sugar cane burning on
aerosol soluble ion composition in Southeastern Brazil’, Atmospheric Environment
38(30): 5025–5038.
Andrade, S.J., J. Cristale, F.S. Silva, G.J. Zocolo, and M.R. Marchi (2010), ‘Contribu-
tion of sugar-cane harvesting season to atmospheric contamination by polycyclic
aromatic hydrocarbons (PAHs) in Araraquara city, Southeast Brazil’, Atmospheric
Environment 44, 2913–2919.
ANFAVEA (2010), ‘Brazilian automotive production’, Historical data, Associac¸ ˜
ao
Nacional dos Fabricantes de Ve´
ıculos Automotores S˜
ao Paulo, Brazil.
Arbex, M., G. B¨
ohm, P. Saldiva, III, A.P.G. Conceic¸˜
ao and A. Braga (2000), ‘Assess-
ment of the effects of sugar cane plantation burning on daily counts of inhalation
therapy’, Journal of the Air and Waste Management Association 50: 1745–1749.
Arbex, M.A., L.C. Martins, R.C. de Oliveira, L.A.A. Pereira, F.F. Arbex, J.E.D.
Canc¸ado, P.H.N. Saldiva, and A.L.F. Braga (2007), ‘Air pollution from biomass
burning and asthma hospital admissions in a sugar cane plantation area in Brazil’,
Journal of Epidemiology and Community Health 61(5): 395–400.
B¨
orjesson, P. (2009), ‘Good or bad bioethanol from a greenhouse gas perspective –
what determines this?’, Applied Energy 86(5): 589–594.
Braga, A.L., G.M. Conceic¸ ˜
ao,L.A.Pereira,H.S.Kishi,J.C.Pereira,M.F.Andrade,
F.L. Gonc¸alves, P.H. Saldiva, and M.R. Latorre (1999), ‘Air pollution and pedi-
atric respiratory hospital admissions in S˜
ao Paulo, Brazil’, Journal of Environmental
Medicine 1(2): 95–102.
Canc¸ado, J.E.D., P.H.N. Saldiva, L.A.A. Pereira, L.B.L.S. Lara, P. Artaxo, L.A. Mar-
tinelli, M.A. Arbex, A. Zanobetti, and A.L.F. Braga (2006), ‘The impact of sugar
cane burning emissions on the respiratory system of children and the elderly.’,
Environmental Health Perspectives 114(5): 725–729.
CIIAGRO (2010), Centro Integrado de Informac¸ ˜
oes Agrometeorol´
ogicas (Center
of Agrometeorological Information), Technical Report, Instituto Agronˆ
omico de
Campinas, S˜
ao Paulo, Brazil.
Crutzen, P.J., A.R. Mosier, K.A. Smith, and W. Winiwarter (2008), ‘N2O release from
agro-biofuel production negates global warming reduction by replacing fossil
fuels’, Atmospheric Chemistry and Physics 8(2): 389–395.
DATASUS (2010), Informac¸˜
oes de Sa ´
ude, Technical report, Minist´
erio da Sa ´
ude.
EIA (2006), International Energy Outlook, Report, Energy Information Administra-
tion, D.F. Brasilia, Brazil.
EPE (2010), Balanc¸o Energ´
etico Nacional 2010, Technical Report, Empresa de
Pesquisa Energ´
etica, Minist´
erio de Minas e Energia, D.F. Bras´
ılia, Brazil.
Farhat, S., R. Paulo, T. Shimoda, G. Conceic¸˜
ao,C.Lin,A.Braga,M.Warth,and
P. Saldiva (2005), ‘Effect of air pollution on pediatric respiratory emergency room
visits and hospital admissions’, Brazilian Journal of Medical and Biological Research
38(2): 227–235.
Giampietro, M., S. Ulgiati, and D. Pimentel (1997), ‘Feasibility of large-scale bio-
fuel production – does an enlargement of scale change the picture?’, Bioscience
47(9): 587–600.
140 Alexandre C. Nicolella and Walter Belluzzo
Gonc¸alves, F., L. Carvalho, F. Conde, M. Latorre, P. Saldiva, and A. Braga (2005), ‘The
effects of air pollution and meteorological parameters on respiratory morbidity
during the summer in S˜
ao Paulo city’, Environment International 31(3): 343–349.
Gunkel, G., J. Kosmol, M. Sobral, H. Rohn, S. Montenegro, and J. Aureliano (2007),
‘Sugar cane industry as a source of water pollution – case study on the situation in
Ipojuca river, Pernambuco, Brazil’, Water Air and Soil Pollution 180(1–4): 261–269.
IBGE (2010a), Censo Agropecu´
ario, Historical data, Instituto Brasileiro de Geografia
e Estat´
ıstica, Rio de Janeiro, Brazil.
IBGE (2010b), Pesquisa Agropecu´
ario Municipal, Historical data, Instituto Brasileiro
de Geografia e Estat´
ıstica, Rio de Janeiro, Brazil.
IPEADATA (2010), ‘´
Indice Nacional de Prec¸os ao Consumidor’, Technical Report,
Instituto de Pesquisa Econˆ
omica Aplicada.
Lara, L., P. Artaxo, L. Martinelli, P. Camargo, R. Victoria, and E. Ferraz (2005),
‘Properties of aerosols from sugar-cane burning emissions in Southeastern Brazil’,
Atmospheric Environment 39(26): 4627–4637.
Mazzoli-Rocha, F., C.B. Magalh˜
aes, O. Malm, P.H.N. Saldiva, W.A. Zin, and D.S.
Faffe (2008), ‘Comparative respiratory toxicity of particles produced by traffic and
sugar cane burning’, Environmental Research 108(1): 35–41.
Moreira, J.R. (2000), ‘Sugarcane for energy – recent results and progress in Brazil’,
Energy for Sustainable Development 4(3): 43–54.
Moreira, J.R. and J. Goldemberg (1999), ‘The alcohol program’, Energy Policy
27(4): 229–245.
Oliveira, M.E. D.d., B.E. Vaughan, and E.J. Rykiel Jr. (2005), ‘Ethanol as fuel: energy,
carbon dioxide balances, and ecological footprint.’, Bioscience 55(7): 593–602.
Ribeiro, H. (2008), ‘Queimadas de cana-de-ac¸ ´
ucar no Brasil: efeitos `
asa
´
ude respi-
rat´
oria’, Revista de Sa´ude P ´ublica 42: 370–376.
Roseiro, M.N.V. and M.M.T. Angela (2006), ‘Morbidade por problemas respirat´
orios
em Ribeir˜
ao Preto (SP) de 1995 a 2001, segundo indicadores ambientais, sociais e
econˆ
omicos’, Revista Paulista de Pediatria 24(2): 163–170.
SEADE (2010), Informac¸˜
oes dos munic´
ıpios paulistas – imp, Technical report,
Fundac¸˜
ao Sistema Estadual de An´
alise de Dados, S˜
ao Paulo, Brazil.
Sicard, P., A. Mangin, P. Hebel, and P. Mall´
ea (2010), ‘Detection and estimation
trends linked to air quality and mortality on French Riviera over the 1990–2005
period’, Science of the Total Environment 408(8): 1943–1950.
SMA (2010), Colheita mecanizada – projeto CANASAT, Technical Report, Secretaria
do Meio Ambiente do Estado de S˜
ao Paulo, S˜
ao Paulo, Brazil.
Uriarte, M., C.B. Yackulic, T. Cooper, D. Flynn, M. Cortes, T. Crk, G. Cullman,
M. McGinty, and J. Sircely (2009), ‘Expansion of sugarcane production in S˜
ao
Paulo, Brazil: implications for fire occurrence and respiratory health’, Agriculture,
Ecosystems & Environment 132(1–2): 48–56.
... Despite all this, the effect of the burning practice before the harvesting season, has been, and still is, the subject of several academic and health debates. Some studies reveal that there is no significant effect from the sugarcane burning practice on the local or nearby population (Jose Goldemberg, 2007), while other references indicate that there are negative impacts, which manifest as respiratory diseases in children and elderly people that receive treatment in local hospitals (Nicolella & Belluzzo, 2011). There are some ongoing studies regarding the potential hazardous effect of the sugarcane burning practice on human health, but research and additional monitory controls are required to obtain conclusive results on the possible carcinogenic outcomes from such procedures. ...
... Summer smog and respiratory diseases can be caused by a frequent practice in sugarcane cultivation which is the pre-harvesting burning process (in the case of palm oil this phenomenon come from the production and use stages). For ET-OH there is no consensus on the net effect of such practice on human health: while some studies show indicate that there is no significant effect on the local population (Jose Goldemberg, 2007), whereas some other studies unveil negative effects on children and elderly people, due to respiratory diseases (Nicolella & Belluzzo, 2011). Within this study the PM effect due to pre-harvesting burning task is assumed in low density areas in terms of population. ...
... Additionally, the pre-harvesting burning has a significant impact on the air quality and it might affect the quality of the environment as a whole and of course human health (by smog and particulate matter). Is not conclusive the effect on human health on the nearby population: whereas some authors argue that there is no evidence of harm on the locals (Jose Goldemberg, 2007), some other authors disagree and explain that such practice affects in major extent to elderly people and children due to respiratory diseases (Nicolella & Belluzzo, 2011). ...
Article
Full-text available
Bioenergy has emerged as a potentially sustainable alternative to the use of fossil fuels for transport and industrial uses. Developing nations, such as Colombia, can seize the advantages of modernizing rural areas by using cleaner energy and having more economic opportunities with bioenergy initiatives, provided the trade-offs between fiber, food, feed and fuel can be managed. This Thesis examines the bioenergy program now under way in Colombia, where comparative advantages (shared with other tropical countries) in production of sugar cane and palm oil are being built on. While the technologies associated with use of these feedstocks are well known, nevertheless their scaling up in a country like Colombia poses considerable environmenral, social, economic and business challenges.
... Air pollution, occurring both outdoors and indoors, is widely recognized as a significant threat to human health (Guarnieri and Balmes 2014;Lelieveld et al. 2018;Prüss-Ustün et al. 2015). Several recent studies find air pollution to harm respiratory health (Ghosh and Mukherji 2014;Kim 2021;Komisarow and Pakhtigian 2021;Morello 2023;Nicolella and Belluzzo 2014;Rocha and Sant'Anna 2022), birth outcomes (DeCicca and Malak 2020; Long et al. 2021;Rangel and Vogl 2019), and longevity (Blackman et al. 2023;Deryugina et al. 2019;Jayachandran 2009;Liu and Zhang 2023). This growing body of evidence suggests that reducing air pollution exposures can improve health in the short-and long-terms, yet it also highlights heterogeneities in impact stemming from differential exposures and vulnerabilities (Morello 2023;Rocha and Sant'Anna 2022) as well as non-linearities in effects (Chay and Greenstone 2003;Shen et al. 2017;Yang et al. 2022). ...
... One branch of this literature focuses on outdoor air pollution exposure and health outcomes (Frankenberg et al. 2005;Ghosh and Mukherji 2014;Jayachandran 2009;Nicolella and Belluzzo 2014;Rangel and Vogl 2019;Tan-Soo and Pattanayak 2019). Many of these studies rely on quasi-experimental data and use plausibly exogenous outdoor air pollution shocks to justify causal claims. ...
Article
Full-text available
Burning tropical forests to establish lucrative agricultural crops ignores potentially important health externalities of the resulting air pollution. These health externalities are often poorly understood, especially if other environmental hazards, such as indoor pollution, are not taken into account. Given the potential for joint, contemporaneous harms, we estimate the impacts of outdoor and indoor air pollution on respiratory health in Indonesia. To address the endogeneity of air pollution exposure, we use panel fixed effects estimation and instrument for outdoor pollution using upwind forest fire intensity. We find that outdoor air pollution exposure reduces lung capacity and decreases overall health status. Subgroup analysis reveals that these impacts are higher among the youngest and oldest individuals in our sample. Critically, we find suggestive evidence that outdoor air pollution exposure is more harmful to the health of individuals living in households that use clean cooking fuels. Thus, policies aimed at reducing environmental health harms are not substitutable—that is, reductions in both indoor and outdoor air pollution exposures are necessary for achieving health targets.
... Burning of sugarcane fields before harvesting is a common practice in many developing countries to reduce the labor cost of leaf removal. However, this practice has also caused significant air pollution, especially particulate matters with diameter less than 2.5 μm (PM 2.5) and carbon dioxide emission [2][3][4]. To reduce such a harmful effect, conversion of unwanted sugarcane leaves into value-added products is necessary [5]. ...
Article
Full-text available
In an attempt to enhance the value of sugarcane leaf, xylan was extracted and used for xylooligosaccharide (XO) production via enzymatic hydrolysis using xylanase from the black yeast Aureobasidium pullulans. The xylan was extracted from sugarcane leaf using alkali extraction according to the response surface methodology. The highest xylan yield (99.42 ± 4.05 % recovery) was obtained using 14.32 % (w/v) NaOH, 13.25:1 liquid: solid ratio, at 121 °C and 15 lb.in² for 32 min. Sugar composition and FTIR spectrum analyses confirmed its structure as arabinoxylan. The extracted arabinoxylan had a relatively high molecular weight compared to previous studies. Crude endoxylanase from A. pullulans NRRL 58523 was selected for enzymatic hydrolysis of the xylan. The enzyme hydrolyzed well at 50 °C, pH 4.0 and was relatively stable under this condition (87.38 ± 1.26 % of the activity remained after 60 h). XOs, especially xylobiose and xylotriose, were obtained at the maximum yield of 237.51 ± 17.69 mg/g xylan via endoxylanase hydrolysis under the optimum conditions (50 °C, pH 4.0, 65.31 U/g xylan, 53 h). XOs exhibited species-specific prebiotic activity toward three strains of Lactobacillus spp. but not toward Bifidobacterium spp.
... Economic analysis of the health effects of air pollution from biomass burning has grown more recently as biomass smoke has become an increasingly important source of pollution due to effective regulation of industrial sources and the effects of climate change on wildfires (Burke et al. 2021). For example, pollution from biomass burning may be related to agricultural fires, which have been shown to have negative consequences for health in India (Pullabhotla and Souza 2022;Singh et al. 2019), China (Lai et al. 2022), andBrazil (Carrillo et al. 2019;Nicolella and Belluzzo 2015;Rangel and Vogl 2019). Evidence on the health impacts of pollution from forest fires has largely focused on Indonesia (Jayachandran 2009; Rosales-Rueda and Triyana 2019; Sheldon and Sankaran 2017; Tan-Soo and Pattanayak 2019) and the US (DeFlorio- Barker et al. 2019;Moeltner et al. 2013;Reid et al. 2019), showing poorer health outcomes at birth and in early childhood for those exposed to smoke in utero, with persistent long term effects; and contemporaneous respiratory health consequences for adults, particularly elderly adults. ...
Article
Full-text available
There is growing recognition of the connection between ecosystem conservation and human health. For example, protection of tropical forests can affect the spread of infectious diseases, water quality, and dietary diversity, while forest loss can have important consequences for respiratory health due to the use of fire for converting land to alternative uses in many countries. Studies demonstrating links between ecosystems and health often conclude with recommendations to expand policies that protect natural ecosystems. However, there is little empirical evidence on the extent to which conservation policies actually deliver health benefits when they are implemented in real contexts. We estimate the effects of protected areas (PAs), the dominant type of conservation policy, on hospitalizations for respiratory illness in the Brazilian Amazon biome. We find that doubling upwind PAs reduces PM2.5 by 10% and respiratory hospitalizations by 7% in the months of most active biomass burning. Brazil has an extensive network of PAs, but investments in management and enforcement have declined in recent years. Forest fires have increased dramatically over the same period. We estimate that the value of the health benefits exceed current average expenditures on PA management for the 1/3 of PAs with the largest local populations, although not for PAs in more remote locations. Our findings highlight how quantifying the contributions to the wellbeing of local populations can support conservation objectives, even if global environmental benefits are not a high priority for decision makers.
... As mentioned above, many farmers overlook the potential uses of crop residues, and a large amount is burned annually, with varying intensities across countries. 2 For instance, in Brazil, sugarcane preharvest burning is widespread, raising serious environmental and health concerns (Nicolella & Belluzzo 2015). Residue burning is not widely practiced in Sub-Saharan Africa, except for cotton and sugarcane (Scholes et al. 2011). ...
Article
Full-text available
More than five billion metric tons of agricultural residues are produced annually worldwide. Despite having multiple uses and significant potential to augment crop and livestock production, a large share of crop residues is burned, especially in Asian countries. This unsustainable practice causes tremendous air pollution and health hazards while restricting soil nutrient recycling. In this review, we examine the economic rationale for unsustainable residue management. The sustainability of residue utilization is determined by several economic factors, such as local demand for and quantity of residue production, development and dissemination of technologies to absorb excess residue, and market and policy instruments to internalize the social costs of residue burning. The intervention strategy to ensure sustainable residue management depends on public awareness of the private and societal costs of open residue burning.
... It clearly showed that the juice quality of burnt cane deteriorated (increased reducing sugars) more than that in the green harvested canes. The results are in agreement with the findings of Nicolella and Belluzzo (2015), who reported that burnt deteriorated faster than the green cane after harvesting of sugarcane. ...
Article
Full-text available
The study aimed to determine the effects of cane storage periods on pre-harvested burnt and unburnt green harvested canes on quality characteristics and sugar yield of sugarcane varieties at Finchaa Sugar Factory, Ethiopia. Experiments were conducted on a split–split plot design with two varieties, N 14 and N Co 334 as main plots, types of canes burnt and unburnt green harvested canes as sub-plots, and eight cane storage periods (0, 24, 48, 72, 96, 120, 144, and 168 h) as sub–subplots with three replications. All quality parameters were investigated using the standard procedure. The results showed that the loss in cane weight increased and pol %, purity % juice, estimated recoverable sugar, and sugar yield decreased significantly with increasing periods of storage. The results also indicated that losses were more in burnt harvested canes than the unburnt canes over the storage period. Brix and reducing sugar in juice increased with the storage periods and these were more in burnt canes than unburnt harvested canes. The loss in cane weight, pol, juice purity, recoverable sugar, and sugar yield was high in NCo334 variety at each storage period. The pol and purity % juice, recoverable sugar % cane, and sugar yield were more in burnt and unburnt harvested canes in both varieties at storage periods. The findings suggested that the N 14 cane variety of burnt harvested canes and green canes should be crushed within 24 and 48 h to obtain better sugar quality than N Co 334 variety.
Article
Full-text available
Sugarcane is the most widely cultivated crop in the world, with equatorial developing nations performing most of this agriculture. Burning sugarcane is a common practice to facilitate harvest, producing extremely high volumes of respirable particulate matter in the process. These emissions are known to have deleterious effects on agricultural workers and nearby communities, but the extent of this exposure and potential toxicity remain poorly characterized. As the epidemicof chronic kidney disease of an unknown etiology (CKDu) and its associated mortality continue to increase along with respiratory distress, there is an urgent need to investigate the causes, determine viable interventions to mitigate disease andimprove outcomes for groups experiencing disproportionate impact. The goal of this review is to establish the state of available literature, summarize what is known in terms of human health risk, and provide recommendations for what areas should be prioritized in research.
Article
Full-text available
Background: Ethiopia's sugar factories are growing by creating job opportunities for thousands of workers with varying educational, professional and socioeconomic backgrounds. These sugar factories are a source of several hazards that severely harm the workers' health. In this context, there is inconclusive evidence on the level of bagasse dust exposure and chronic respiratory health symptoms. This study aimed to assess the degree of bagasse dust exposure and chronic respiratory health symptoms. Methods: In this longitudinal study, five workstations were selected for dust sampling. A stratified random sampling technique was used to select 1043 participants. We measured the dust intensity using a calibrated handheld real-time dust monitor once a month for 5 months, totalling 50 dust samples. Chronic respiratory symptoms were assessed using the American Thoracic Society's respiratory symptoms questionnaire. Results: A 1 hour time-weighted average of bagasse dust intensity in the boiler, power turbine and evaporation plant was 8.93 mg/m3, 8.88 mg/m3 and 8.68 mg/m3, respectively. This corresponded to an exposure level to bagasse dust of 85.52% (95% CI 83.2% to 87.6%). The level of chronic respiratory health symptoms was 60.6% (95% CI 59.2% to 61.9%). The most common respiratory symptoms were wheezing (96.8%), coughing (89.7%) and breathlessness (80.9%). The identified risk factors were lack of dust control technology (β= 0.64, 95% CI 0.53 to 0.75), not practising wet spray (β = 0.27, 95% CI 0.21 to 0.41) and not wearing proper respiratory protection devices (β = 0.12, 95% CI 0.30 to 0.56). Conclusions: Bagasse dust exposure and respiratory health abnormalities were worrying concerns. The absence of dust control technologies and no practice of wet spraying elevated the level of exposure. Not wearing proper respiratory protection gear increased the odds of having respiratory abnormalities. Hence, the use of mechanical solutions to stop dust emissions at their sources and the wearing of proper respiratory protection gear are highly advised.
Article
Full-text available
Bioenergy in general, and biofuels in particular, have come up to the renewable energy stage with some peculiar strength, overall in terms of alternatives for transportation. Some of the drivers behind this option are shared on a global scale, such as the reduction of GHG’s emissions, and enhancement of energy security conditions. Some others have a more local nature, like a diversification of markets for agricultural commodities, dynamization of rural areas, improvement of micro and macro-economic indicators (for instance, income of the rural poor and national balance of payment), among others.
Article
Full-text available
The major contributor to global warming is considered to be the high levels of greenhouse gas emissions, especially carbon dioxide (CO2), caused by the burning of fossil fuel. Thus, to mitigate CO2 emissions, renewable energy sources such as ethanol have been seen as a promising alternative to fossil fuel consumption. Brazil was the world's first nation to run a large-scale program for using ethanol as fuel. Eventually, the United States also developed large-scale production of ethanol. In this study, we compare the benefits and environmental impacts of ethanol fuel, in Brazil and in the United States, using the ecological footprint tool developed by Wackernagel and Rees. We applied the STELLA model to gauge possible outcomes as a function of variations in the ethanol production scenario.
Article
Full-text available
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.
Article
Full-text available
The article aimed to update scientific literature information about respiratory health effects caused by sugarcane burning, considering the expansion of sugarcane plantations in Brazil and in the state of São Paulo. Articles published between 1996 and 2006, which deal with the health effects of sugarcane burning and/or air pollutants originating from this burning, were discussed. These studies suggest that part of the population – especially the elderly, children and asthmatics – suffers health effects of sugarcane burning. As a result, these people require health care, thus affecting health services and their families.
Article
Full-text available
The major contributor to global warming is considered to be the high levels of greenhouse gas emissions, especially carbon dioxide (CO2), caused by the burning of fossil fuel. Thus, to mitigate CO2 emissions, renewable energy sources such as ethanol have been seen as a promising alternative to fossil fuel consumption. Brazil was the world's first nation to run a large-scale program for using ethanol as fuel. Eventually, the United States also developed large-scale production of ethanol. In this study, we compare the benefits and environmental impacts of ethanol fuel, in Brazil and in the United States, using the ecological footprint tool developed by Wackernagel and Rees. We applied the STELLA model to gauge possible outcomes as a function of variations in the ethanol production scenario.
Article
Full-text available
Brazil's rise to be the world's preeminent bioenergy pro- ducer provides three important lessons. The first lesson is about the complex task for developing countries balancing government intervention with market forces as they try to develop an industry. The second is how critical research and development (R&D) is for lowering costs to allow for market entry of an infant industry. The third is about the new challenges for bioenergy as it increasingly competes with the food industry for the same raw materials. The Industry Increases in petroleum prices and demand are creating pressure to develop new sources of renewable energies. Biofuel will represent 30% of the global energy used by 2020 compared with only 2% today (International Energy Agency, 2005). In 2004, the global ethanol market was US$30-40 billion, of which $4 billion involved interna- tional trade. Brazil, China, India, Malaysia and South Africa, the United States (US), and the European Union (EU) are important players in the burgeoning global mar- ket. Brazil is one of the world's most competitive biofuels producers because of its comparative advantage in produc- ing ethanol and soybeans. The US, the 2 nd leading ethanol producer in the world, has variable costs of production of corn-based ethanol of US$0.96 per gallon. Fixed costs range from US$1.05 to US$3.00 per gallon. While in Bra- zil the total cost of production was approximately US$1.10 per gallon during the 2005 crop year, with vari- able costs of US$.89 per gallon and fixed costs of US$.21 per gallon. In early 2006, the wholesale price paid to the mills for anhydrous ethanol was US$2.05 per gallon, while the retail price at the time for ethanol-gasoline blends was US$3.41 (including taxes). Total world ethanol production (all grades) in 2005 was 12.2 billion of gallons, with 70% of this total pro- duced by the US and Brazil (Figure 1). Other significant producers are China, the EU, and India. Production in the
Article
Full-text available
The Ipojuca river, a river in Northeast Brazil extending from west to east, is situated in the semiarid interior and rainy coastal zone of Pernambuco state. The river basin covers an area of 3,514km2, with a total river length of 215km. The water flow regime is characterized by the annual change from dry to rainy season as well as periodically occurring dry cycles. Mean flow rates vary between 2 and 35m3 s−1. The dominant impacts on water quality of the river are domestic sewage input in the upper catchment and sugar cane cultivation and processing in the lower catchment. Long-term monitoring data used to demonstrate the impact of sewage discharge on the river’s self-purification capacity indicated a severe change to the worse in the lower course. One reason is the use of stillage (wastewater from cane processing) for fertilization and irrigation (fertigation). Pathways of contamination have been identified by evaluating cultivation and processing techniques of a bio-alcohol factory with annexed sugar cane cultivation. The river’s main ecological problems are water heating, acidification, increased turbidity, oxygen imbalance, and increased coliform bacteria levels. Precipitation-related wash-off and wash-out causes significant contamination within one to two days after rainfall. The increase in coliform bacteria is a consequence of secondary contamination. To determine the impact of the sugar cane industry on the river, the Brazilian bio-alcohol program must be subject to a critical evaluation. Environmentally friendly sugar cane cultivation methods need to be developed, and sugar cane factories must use waste-reducing technologies and water cycling processes in order to protect the region’s water resources.
Article
The rational for the launching of the Alcohol Program from sugarcane in Brazil in the mid-1970’s is described as an answer to the first “oil crisis” as well as a solution to the problem of the fluctuating sugar prices in the international market. The technical characteristics of ethanol as a fuel are given as well as a discussion of the evolution of the cost of production, environmental and social consequences. Regarding costs, ethanol production was close to 100 dollars a barrel in the initial stages of the Program in 1980 falling rapidly due to economies of scale and technological progress to half that value in 1990, followed by a slower decline in recent years. Considering the hard currency saved by avoiding oil importation through the significant displacement of gasoline by ethanol and the decrease in the amount of external debt that the displaced oil importation was able to provide it is possible to demonstrate that the Alcohol Program has been an efficient way of exchanging dollar debt by national currency subsidies which are paid by the liquid fossil fuel users. Even with this economic gains for society, the continuity of the Program is difficult to maintain. Two solutions to this problem are discussed: internal expansion of the use of ethanol and exports to industrialized countries where it could be used as an octane enhancer. The main attractiveness of the Program – the reduction of CO2 emissions as compared to fossil fuels – is stressed, mainly as a solution for industrialized countries to fulfill their commitments with the United Nations Framework Climate Change Convention.
Article
Epidemiological studies have shown that high levels of fine particulate matter (PM) are correlated with adverse human health effects. Approximately one-third of PM emissions in Canada originate from forest fires. However, air quality concerns are not typically included in resource allocation decisions in fire management. In this paper we examine the economic magnitude of these health concerns and compare them to other costs of forest fires using the 2001 fire in Chisholm, Alberta, as a case study. We construct an empirical air dispersion model to estimate the concentration of PM arising from the fire. Benefit transfer methods were used to determine the health impacts associated with elevated PM from the fire and to value these impacts. The economic impacts appear to be substantial, second only to timber losses. The approaches used in this case study can be extended to construct a map that identifies the values at risk from health effects. The use of monetary values of these impacts helps in comparison and aggregation of the values at risk.
Article
Ethanol is being used as a fuel in the transportation sector in Brazil and in a few other countries. It started to be used in significant amounts in Brazil in a blend with gasoline in 1975 and as a neat fuel in automobiles in 1980. The first use continues and all gasoline sold in the country contains 24% ethanol. The second use showed a rapid evolution, covering more than 90% of the new cars produced in 1989; after a shortage of supply in that year the number of new neat ethanol automobiles decreased very fast, reaching less than 1000 cars by 1997–98. By 1999 a modest reverse of this trend had occurred. Considering both types of use, ethanol demand has stabilized since 1995, but at present some decline in demand is foreseen for new neat ethanol cars for several years in the first decade of the 2000s. The government has an interest in maintaining the use of this alternative fuel because of several merits, e.g., the large number of jobs created in the field and the significant amount of hard currency saved on oil imports. On the other hand, most of the subsidies provided to ethanol producers have been removed because they have been in existence for a long time. Some subsidies are still in force, such as the over-price paid to producers in some states where it is understood that they are not capable of competing with the major producers of the South/Southeast region. There is enough evidence that through the “learning by doing” process ethanol fuel producers are learning how to bring down costs and they already have the ability of competing with gasoline, which is heavily taxed to cover the sales of some petroleum products at prices below production costs. Technology has improved and further progress may be made, allowing a reduction in the ethanol fuel production cost. Nevertheless, oil prices for several years up to 1999 also declined, making the competition between gasoline and ethanol difficult. By 1999, with the rebound in oil prices, the federal government reinvigorated its interest in the ethanol market through some political initiatives. Such evidence of interest and the present increase in oil price may offer the opportunity required to restart large scale production of neat ethanol automobiles. The electricity market is opening up space for the use of large amounts of biomass residues, obtained during ethanol processing, as a fuel in steam boilers. The interest is growing and more modern technologies can be used, allowing the production of up to 6000 MW of electricity using the sugarcane residues, bagasse and barbojo. Another technology which is under consideration is the development of suitable re-formers to allow the use of ethanol in fuel cell-powered vehicles. Considering the significant success of ethanol as an automobile fuel and for electricity generation this paper tries to make an assessment of the future trend. It is very useful to recognize that during the last 40 years, average sugar prices (the only alternative market for ethanol from sugarcane) have been continuously declining, while average gasoline prices have moved in the opposite direction. This trend helps commercial competition between the two fuels. In addition, the potential economic value of a renewable fuel through carbon emission certificates can impact positively on ethanol. Significant carbon abatement results from the use of ethanol derived from sugarcane due to the very favorable energy balance of the entire process.
Article
In order to investigate the relation between air pollution and child morbidity in São Paulo, a time-series study was carried out. Daily records of hospital admissions for children under 13 years old were obtained at the Health State Secretary, covering 112 hospitals in the period from October 1992 to October 1993. Daily levels of PM10, ozone, SO2, CO and NO2 were obtained from the environmental state agency (CETESB), while both CETESB and the Institute of Astronomy and Geophysics (IAG) of the University of São Paulo provided daily measures of temperature and relative humidity. Daily counts of child respiratory admissions (RESP) were considered as the dependent variable of pollutants in regression models, controlled for months of the year, days of the week, weather factors, and the daily number of non respiratory admissions (NORESP). PM10 and ozone were the pollutants that exhibited the most robust association with RESP. The mean levels of PM10observed during the period of study (70 μg m−3) were associated with an increase of 12% in RESP. The association between air pollution and RESP was significant within a time lag between 1 to 7 days and was dose-dependent. This result indicates that air pollution represents a significant pediatric health problem in São Paulo. Copyright © 1999 John Wiley & Sons, Ltd.