Salaries,
incentives and intellectual output of teachers in public universities in
Colombia
Salarios,
incentivos y producción intelectual docente en la universidad pública en
Colombia*
Salários, incentivos e
produção intelectual docente na universidade pública em Colômbia
Jhon Alexánder
Méndez Sayago**
Leonardo
Vera Azaf***
Research article
Date of
reception: 16 oct
2014
Date of
approval: 25 jun
2015
Abstract
In this article
the determinants of the intellectual productivity of professors in public
universities in Colombia are studied, focusing on the effect of salary
incentives on intellectual productivity and the resulting salary. Based on
information from the teaching staff of the Universidad del
Valle between the years 2004-2012, econometric models of the lifecycle of
university professors were estimated to quantify these effects. An important
and significant effect of the present value of the salary point was found, but
no positive effect was found on wages due to intellectual output. The effect of
incentives according to the age of the researcher and the impact of percentage
increases in the value of the salary were also calculated.
Keywords: intellectual output, life cycle model, university
professor, Tobit model, incentive.
JEL: C24, J24
Resumen
En este artículo se estudian los determinantes de la
productividad intelectual de los profesores de la universidad pública en
Colombia, enfocándose en el efecto del incentivo salarial por productividad
intelectual y el efecto salario. A partir de información del cuerpo de
profesores de la Universidad del Valle, comprendida entre los años 2004-2012,
se estimaron modelos econométricos de ciclo de vida del profesor universitario
para cuantificar dichos efectos. Se encontró un efecto importante y
significativo del valor presente del punto salarial, pero no se halló efecto
positivo del salario sobre la producción intelectual. También se calculó el
efecto del incentivo en función de la edad del investigador y el impacto de
incrementos porcentuales en el valor del punto salarial.
Palabras clave: producción intelectual, modelo de ciclo de vida,
profesor universitario, modelo Tobit, incentivo.
Resumo
Este artigo
discute os determinantes da produtividade intelectual de professores de
universidades públicas na Colômbia, concentrando-se sobre o efeito da produtividade
intelectual salário de incentivo e efeito salário são estudados. A partir de
informações do corpo docente da Universidad del Valle, entre os anos de
2004-2012, professor econométrico modelos de ciclo de vida universitária para
quantificar esses efeitos foram estimados. Um efeito importante e significativo
do valor presente do ponto de salário foi encontrado, mas nenhum efeito
positivo dos salários foi encontrado na produção intelectual. O efeito do incentivo,
dependendo da idade do investigador e do impacto de percentagens de aumento no
valor do ponto de salário também foi calculada.
Palavras-chave: produção
intelectual, modelo de ciclo de vida, professor universidade, modelo Tobit, incentivo.
INTRODUCTION
As indicated by Baccini, Barabesi, Cioni, and Pisani (2014) and Obembe (2012), the study of scientific productivity and of
the factors associated with it, has been a topic that has attracted the
attention of researchers, particularly in the last 30 years.
Stephan (2010)
expresses three reasons for which the economy of the sciences has been gaining
ground in the last years. Firstly, science has been identified as a source of growth,
which has been confirmed by the recent advances in information technologies.
This situation has contributed significantly to the growth of the service
sectors in recent years. Advances in medical research have also led to an increase
in work and life expectancy. The second reason is related to the topic of
appropriation of knowledge, that is to say, once knowledge is produced and made
public, no one can be excluded from its use. Therefore, the failure of
economies when trying to produce the optimal amount of knowledge considering it
a public good is a topic of importance for the economists. Finally, the public
nature of research and the spillover inherent in a system of such
characteristics is fundamental for the concept of the theory of endogenous
growth (Archibugi & Coco, 2004).
Perozo, Arteaga and Fuenmayor (2008)
argue that research has become the process that contributes the most to the
development of countries, which explains why powerful countries invest high percentages
of their gross domestic product (GDP) in research and development, achieving
great scientific and technological advances. This situation leads to the affirmation
that research is recognized globally as a pillar of the acquisition of new
knowledge and technologies. This complex activity involves companies,
institutions, national universities, and international collaboration networks
so as to obtain better results.
The current
demands of society also force professors, from diverse scientific standpoints,
to seek concrete solutions to the problems that the educational context in
which they develop requires. It is from there that the idea arises that
universities should be understood as centers of intellectual output, taking into
account that, in the academic context, it is research that generates knowledge.
In a knowledge society, the creation of new knowledge is indispensable, and it
is the university professor who is called to produce and revise it. In fact,
the results of investigations help to improve the praxis of the teacher in a
permanent way and, therefore, research represents a primary activity in
universities.
Facing these
changes, Colombian universities, especially public universities and the most
well-known private universities, that traditionally focused on teaching,
changed their perspective and developed policies that make research one of
their priorities, offering incentives to publications and expecting that
through them, teachers would generate knowledge and circulate it in national
and international indexed journals, research books, or take part in conferences
and academic events (Guzmán & Trujillo, 2011).
In this
incentive policy, basically two strategies can be identified. The first is
implementing starting salaries which are relatively low that increase based on
intellectual output. In the second, relatively high starting salaries are
offered with premiums for intellectual output during a limited period,
generally, of one year.
One of the main
problems that public universities in Colombia face in the recruitment of new
professors are the legal restrictions they have in setting a salary that is
attractive to the most talented candidates in the market[1]. Nevertheless,
public universities also offer advantages that motivate professors with
research abilities to join public universities: 1) they do not demand exclusive
contracts, 2) depending on labor seniority, experience and, above all,
productivity, a public university professor may earn as much as a private
university professor.
The problem is
that once a teacher is hired, given their low salary, they could opt for other
activities that generate income to the detriment of the time dedicated to
research, which would affect intellectual output, which is one of the aims of
the university management as that brings prestige to the institution.
This article
intends to establish which of the two alternatives is more effective as an
instrument for stimulating the intellectual output of teachers. To achieve this
aim, the article researches the determinants of intellectual productivity of professors
in public universities in Colombia, focusing on the effect of incentive pay for
intellectual productivity and the resulting salary.
In order to
understand the relationship between salary, incentive pay, and intellectual output,
the article presents an adaptation in discrete time of the Levin and Stephan
life cycle model (1991). The innovation is that the future salary of the researcher
professor depends on publications during their life cycle, and not on their
prestige.
Afterwards, from
the information provided by the body of professors from the Universidad del Valle between 2004-2012, the life cycle model of a
university professor was estimated econometrically so as to quantify the
effects of salary and incentive pay on intellectual output.
The article is
organized as follows: in the first section, a revision of the state of the
situation is presented, which gives an account of the main determinants of
scientific production and makes a revision of the theoretical life cycle models
which attempt to explain the dynamic of the productivity of a researcher
through time. In the second section, the contractual relationship between the
public university and the university professor is analyzed as from agency
theory, with emphasis on incentives as a stimulus for intellectual output. In
section three, a life cycle model for university professors in public
universities in Colombia is developed, which unveils the impact of salary and
incentive pay on their intellectual output. In section four, data and
descriptive statistics are presented. In section five, an econometric
estimation of the life cycle model is realized. In section six, the results are
analyzed so as to finally present the conclusions.
REVISION OF THE LITERATURE
Print
and Hattie (1997) define scientific production as the group of studies
developed by academics in universities and related contexts during a certain
period of time. For Piedra and Martínez
(2007), scientific production is understood as the materialized part of the
knowledge that was generated, and for this reason, many refer to it as the results
in the form of publications of research works; such as, for example, articles
published in international, local or national journals, books or book chapters,
presentations in conferences, documents, etc.
Others
place scientific production in a broader setting than the mere publication.
According to that point of view, “the theses that were defended, but not yet
published, the works presented in conferences, colloquiums, and symposiums,
classes, laboratory works finished, but not published, including field work;
all the above is scientific production” (Piedra &
Martínez, 2007, p. 3).
Even
recognizing that scientific productivity covers the whole of the research
produced by scientists, its measurable element is the number of publications
produced by a certain author, institution or country. Manjarrés
(2009) presents arguments such as the peer review of Fielden
and Gibbons (1991), the preferences of researchers, such as Crane (1965) and
Merton (1968), the homogenization of Paasi (2005),
and the perception of prestige of the journal of Miller and Serzan
(1984), so as to conclude that the number of articles published by a professor
in international journals is the principle measure of scientific production of
the present day.
In
regard to this, in Colombia, university professors and magazine editors have spoken
against the way of socializing knowledge imposed by the standards of mainstream
science, which leads to the marginalization of Colombian academic output and
scientific publications.
After
discussing the concept of scientific productivity, the state-of-the-art of the different
forms of analysis of scientific productivity can be explored. Baccini et al.
(2014) distinguish two alternative approximations: the first one focuses on the
fundamental laws of frequency distribution of the number of publications (or
quotes), whereas the second one has the objective of identifying the determinants
of intellectual output.
Baccini et al. (2014) also mention that the
first approximation has its origin in the pioneering study of Lotka (1926) on the frequency distribution of the
scientific performance of chemists and physicists. Lokta
concludes that 60% of these individuals make only one contribution throughout
their lives, so that the number of authors with
Furthermore,
Baccini et al.
(2013) quote the arguments of other researchers that attempt to show the
differences among researchers:
The sacred-spark
hypothesis, proposed by Allison and Stewart (1974) and David (1994),
according to which the differences in productivity reflect the unequal and
predetermined abilities of the researchers.
The
Matthew-effect hypothesis, developed
by Merton (1968), which suggests that it is easier for well-known researchers
to publish their works than for lesser
known researchers, despite that the latter have contributions of equivalent
quality.
Cole
and Cole (1973) put forward the cumulative
advantage hypothesis. The idea is that the recognition received at an early
stage of a researcher’s career can be reinforced with time, as this allows them
easier access to the resources necessary for research. This means that any
advantage will be cumulative.
The
second approximation directs its attention to identifying the individual and
collective determinants of scientific productivity. The former refer to the individual
attributes of the researcher (sex, age, field, etc.) and the latter are factors
related to the environment in which the researcher works and the general reward
system of the science. Manjarrés (2009) makes a
distinction between the academic and socioeconomic environment, the latter
being understood as the effect of financing coming from industrial agents and,
in general, the relationship between university and business.
The
objective of this section is to revise the literature related to the
determinants of intellectual productivity by exploring the approaches based mainly
on the study of individual characteristics and the academic context of the
professor, because in Colombia the relationship between public universities and
businesses is very weak. As Gutiérrez and Berrío
(2011) state, the reality of that relationship in Colombia is each to their own.
Individual factors
The
bibliography regarding the determinants of the scientific productivity of
university professors includes, as explanatory variables, the individual
characteristics of the academics, such as age, sex, position within the
institution, and scientific field. The relationship between age and scientific productivity
is the aspect that has most caught the attention of economists and
sociologists, giving origin to the denominated life cycle models of the
researchers.
Although
the results obtained have been diverse, several studies indicate the existence
of a non-linear relationship, in the shape of an inverted U, between the age
and productivity of researchers (Bayer & Dutton, 1977; Cole, 1979;
Zuckerman & Merton, 1972; Weiss & Lillard,
1982; Levin & Stephan, 1991; González & Veloso,
2007). It has also been pointed out that the peak of productivity is
heterogeneous; it varies according to the field of expertise of the professor
(Levin & Stephan, 1989). Lehman (1958, 1960) found that scientists of the
hard sciences reach a peak of productivity sooner than those who belong to
other disciplines. Others have discovered that the productivity curve has two
peaks (Bayer & Dutton, 1977). On the other hand, Allison and Stewart (1974)
noted a positive correlation between inequality in productivity and age groups,
so that, the greater the age, the greater the differences among the
researchers.
The
first studies on the determining factors of scientific productivity were based
on cross-sectional data and attempted to explain the productivity accumulated
during a certain period of time. However, the results are questioned for the
impossibility of distinguishing between the supposed effect of age and that
which would correspond to the generational cohort. Later on, the availability
of longitudinal data allowed for the analysis of the productivity of
researchers of a same cohort with different ages, as they get older, thus separating
the two effects (Manjarrés, 2009).
Rauber and Ursprung
(2008) based on information from German academic economists, found differences
between different cohorts in the relationship between age and intellectual
productivity. In the researchers of greater age, productivity is maintained
throughout their lives, while in the young, a quite pronounced inverted U shape
is observed.
Analyzing
the determinants of productivity observed during the period 1995-2000, of a
sample of 1134 researchers at the Louis Pasteur University in France, Carayol and Matt (2006) found from a Tobit regression that
age negatively affects productivity, but they reject the inverted U shape.
Given
the importance that the age factor has in the explanation of academic
productivity for this research project, this section is dedicated to the
revision of the main economic life cycle models that contribute to
understanding said relationship.
Professional
experience can also be considered to be a factor that affects intellectual
output. Rebne (1990) and Goodwin & Sauer (1995)
concluded that the maximum performance of a researcher occurs within the first
ten years of work and tends to decrease after 25 or 30 years of academic
activity.
It
has also been indicated that the position the professor occupies within the
institution can be a determining factor in explaining their scientific
productivity. Knorr, Mittermeier, Aichholzer
and Waller (1979) found that when the effect exerted by carrying out
administrative chores is controlled, age ceases to be a significant factor.
Many
studies have shown that intellectual productivity tends to increase depending
on the position of the individual in the academic hierarchy. Manjarrés (2009) cites Cole and Cole (1973), Long (1978),
and Carayol & Matt (2006) who discovered that the
professors with higher rank within an institution, for example full-time
teachers, have greater productivity than teachers of lower rank (junior or
assistant).
A
study by Aksnes, Rorstad, Piro and Sivertsen (2011)
indicated that full-time teachers are the most productive. On average, male
professors publish 9.5 publications in a period of four years, followed by
associate professors with 4.8 publications and doctors 4.5 publications,
whereas doctorate students have the lowest productivity (2.9
publications). With respect to the
differences, Kyvik (1991) observed that professors,
being leaders of research groups, appear in all publications whereas doctorate
students only appear in the ones they are the authors of.
It
should be taken into consideration that the incorporation of variables, such as
academic rank or professional experience, may generate multicollinearity
problems with the age variable, modifying its explanatory capacity or altering
the direction of its influence (Carayol & Matt,
2006).
Another
personal attribute associated with scientific productivity is sex. Many studies
have found great differences between the sexes as regards intellectual
productivity in favor of men (Aksnes et al., 2011b; Kyvik & Teigen, 1996; Cole
& Zuckerman, 1984). For example, Aksnes et al. (2011b) discovered that in almost
all age groups, men are more productive than women. Female scientists tend to publish
generally between 20%-40% less than their male colleagues. In order to explain
said difference, investigations have been carried out that include the marital
status variable to try to explain intellectual productivity. The hypothesis is
that married female researchers could be less productive given their domestic
responsibilities. However, the evidence regarding this point is ambiguous (Reskin, 1978; Astin & Bayer,
1979).
Field
of study is another of the classic factors that can determine scientific
productivity. Dundar and Lewis (1998) found
significant differences in the average number of articles published by
academics in universities in the United States according to the academic
discipline they belong to. The authors realized that while a professor
representative of the social sciences published approximately 2.5 articles
between 1988 and 1991, their colleagues in biology published 9 articles during
the same period. However, Dundar and Lewis suggested
that these results reflect the general differences in the publication trends
between different areas rather than authentic differences as regards the levels
of research activity per field.
Institutional factors
Institutional
factors also arise as possible determinants of scientific production. These
factors are mainly related to the institution, the faculty, or the department
where the professor works, or the group in which the scientist develops their
research work According to Manjarrés, (2009, p. 62):
Policies
and institutional objectives are the framework that underlies the evaluation
processes of scientific performance and the foundation of the reward systems in
science. In this sense, the distinctive features among the policies,
objectives, or the emphasis of the university missions may generate different
organizational cultures, which may have an effect on the scientific
productivity of the professor.
In
order to measure the influence of the characteristics of the institution or the
department over intellectual productivity, Manjarrés
(2009) cites Kyvik (1995), Bonaccorsi
& Daraio (2003), Smeby & Try (2005) as authors who have included
in their econometric models variables like department size. Jordan, Meador, and Walters (1989) have taken into
account the public or private nature of the institution, and Creswell (1986)
and Long (1978) have taken the prestige of the department or the institution
into consideration.
It
is supposed that larger organizations can accumulate more resources for
research, which increases productivity. These accumulative effects may also be
accompanied by economies of scale in scientific production. The results are not
conclusive. While some studies find a positive relationship (Dundar & Lewis, 1998), others find little relation
(Cohen, Florida & Goe, 1994) or only in some
disciplines such as the natural sciences (Kyvik,
1995).
The
relation between the number of teaching hours and productivity is not clearly
determined. Whereas in some studies there appears to be a negative relationship
in terms of opportunity cost (Fox, 1992) and in others the relationship is
positive (Dundar & Lewis, 1998; Kyvik & Smeby, 1994).
Estimation techniques and unit of analysis
In
the literature, different estimation methods are employed. When dealing with the
productivity of count data, most of the studies opt for methods, such as that
of Poisson, the negative binomial, and, in the case of proportions with many
zeros in the dependent variable, the Hurdle model is chosen. The MCO method is
also used with the dependent variable in logarithms and the Tobit model.
The
most frequent unit of analysis in the analysis of intellectual productivity is
the individual, but also studies on different levels can be identified, for
example research group, department, faculties, and even territories (Smeby & Try, 2005; Bonaccorsi
& Daraio, 2003; Dundar
& Lewis, 1998).
Porter
and Umbach (2001) demonstrate that the effect of age
in a model that employs multilevel techniques significantly differs from those
offered by an analysis model on an individual level. Smeby
and Try (2005) find that the age of the researchers has a negative effect on
productivity at an individual level, but that the average age in departments
have positive effects in productivity on a group level. The argument used to
defend their results is that older researchers take on tasks that support the
research, such as the organization of seminars or motivation and supervision of
young researchers who generate important externalities. Moreover, as reputation
is based on the publications made and on professional contacts, older
researchers have a greater capacity to attract financing, prestigious
collaborations, and young researchers with better prospects.
SALARIES AND INCENTIVES FOR INTELLECTUAL OUTPUT IN
PUBLIC UNIVERSITIES IN COLOMBIA
A contract is a
set of rules that facilitate the cooperation and exchange carried out by
rational economic agents because it creates the necessary incentives for the
agents to make the transactions. In any economic interaction, two parts can be
identified: the principal and the agent. The principal is the one who hires the
agent to do a job. The employees (agents) commit to use their abilities to do
the tasks that the employer (principal) asks them to do, and the employer
commits to pay the employees and to maintain a favorable work environment (Gorbaneff, 2003).
The
Principal-Agent model allows the analysis of incentive policies related to
research in public universities in Colombia to be undertaken. In the work
contract that the university (principal) and the professors (agents) agree on,
the latter commit to carrying out research, teaching, and extension activities.
The contract includes a determined remuneration, and specifies the level of
effort that is explicit in the academic workload of the professor, in terms of
the hours that they agree to dedicate to research, teaching, and extension
activities. The principal cannot directly observe the actions of the agent
(actual time dedicated to research), so the professor’s fulfillment of the task
is verified through the products that the professor hands in at the end of the
academic period (Carvajalino & Ariza, 2008).
The university
benefits when a professor publishes their intellectual output because this
gives them recognition and status. Nevertheless, there are no effective control
mechanisms that allow the university to directly influence the professor’s level
of effort, as regards the quality of their intellectual output or their effort
to receive recognition for their work. This creates the need for an incentives
scheme that ensures that the professor makes the maximum effort in the
research, so that the results benefit the university as well as the professor.
Decree 1279 of
June 2002 is the norm that determines the salary and the academic incentives of
a professorial career. It established the incentives for a professor’s
intellectual output in public universities in Colombia. According to this norm,
a professor’s salary is the result of multiplying the value of the salary point
by the accumulated salary points that have been granted to the professor.
The value of the
point is determined every year by the national Government and the number of
points depends on the evaluation of the following factors:
·
Category within the professor pay scale
·
University studies
·
Certified experience
·
Academic production
·
Academic-administrative managerial activities
·
Outstanding performance in teaching practices and
extension activities
The categories
defined in the Decree 1279 for the professor pay scale are: auxiliary
professor, assistant professor, associate professor, and tenured professor. To
the degree that a professor ascends in the professor pay scale, they earn more
salary points. A tenured professor can reach a maximum of 96 points. Professors
ascend this scale depending on their intellectual abilities, merits for
academic achievements, certified teaching experience, and intellectual output
in their professional field or their subject of expertise.
The setting of
the remuneration of university professors is also based in their training, and
this is intended to stimulate professors to not stop at a certain level of
education and to reward the investments that teachers make in human capital, in
their intellectual development. For instance, a professor who holds a Master’s
degree may receive 80 salary points for obtaining their doctorate and up to 10
additional points for each post doctorate.
The evaluation
of certified experience in Decree 1279 gives more points when the candidate has
worked in research institutions of science, technology, humanities, or pedagogy
(37 points per year) than when a professor has worked in a university (4 points
per year).
A professor’s
intellectual output is valued in two different ways: through salary points and
with bonuses.
The value of salary points is set in article
10 of Decree 1279: publications in A1 type journals=15 points, in A2 type journals=12
points, in B type journals=8 points, in C type journals=3 points, according to
the indexing system established by Colciencias;
production of cinematographic or phonographic videos of international
importance=12 points, of national impact=8 points; the publication of books as
a result of an investigation=up to 20 points; the publication of essay books=up
to 15 points per book; translation of books=up to 15 points; national and
international awards=up to 15 points; patents= up to 25 points; technical
production=up to 15 points per innovation; for adaptations=8 points; software
production=up to 15 points.
Salary points
for outstanding teaching practice are given taking into consideration the
students’ evaluation of the professor. Outstanding performance in extension
tasks is evaluated based on a document in which the professor supports the
presence of the university before the community, the academic and social
relevance of the service, its complexity and singularity, an evaluation of the
activity by the community or the institution that benefited from the activity.
The annual salary points assigned to outstanding professors in teaching
practices and extension activities are: professor with tenure; up to 5 points,
associate professor; up to 4 points, assistant professor; up to 3 points, and
auxiliary professor; up to 2 points.
It is clear that
Decree 1279 contains an incentive system to foster the intellectual output of
professors in public universities. This incentive system rewards the
professor’s intellectual output with salary points and brings more recognition
to the university. The amount of economic
incentive (EI) corresponds to the
current net value of the salary point.
Guzmán and Trujillo
(2011) conclude that incentives for research have an effect mainly on young
professors, who have recently joined the institution and whose time is partly
devoted to research, for whom the cost of the effort of carrying out the
research should be covered by the research incentive, so as to avoid that they
only partake in teaching activities and instead divide their efforts between
both activities at an affordable cost to society.
Carvajalino and Ariza (2008) developed a survey for professors of the
Universidad Industrial de Santander, in order to capture the perception of the
professors concerning the efficiency of incentives for research and their
general motivation.
The survey
showed that most of the professors consider that the motivation to belong to a
research group lies in the possibility of development and personal academic
recognition (70.73%). Another motivation is to ascend in the professor pay
scale, for which more than half of the professors surveyed gave a high value. 65.85%
of the professors consider that problem solving is a great motivator. And
finally, 24.39% think that the main motivation is to get a pay raise or a
bonus. The survey revealed that the economic incentive is not the main reason
for professors to do research. The possibility of development and personal
recognition as well as an improvement in the professor pay scale are arguments
that support the hypothesis that professors do research because it improves
their status within the university.
In the same
survey, 14.63% of the professors consider that the university incentivizes
research with pay raises, 19.51% of the faculty supposes that incentives are
given through bonuses, 9.76% believe that the incentive comes from the
distribution of the academic workload, 26.83% think that the incentive comes
from academic distinctions, and 34.15% consider that the incentive is given
through study commissions.
The survey
revealed that less than 10% of the faculty find that the allocation of research
hours of the academic workload is an incentive for research, that is to say
that this mechanism of direct regulation for research does not work, according
to the very same agent (professor).
LIFE CYCLE MODEL FOR PROFESSORS OF PUBLIC UNIVERSITIES
IN COLOMBIA
Life
cycle models of scientific researchers are intertemporal optimization models of
the profit or income of the researcher, which relate
the time dedicated to research and, thus, relate their research results with
age. As their stock of knowledge gives them prestige, this depends on time
dedicated to research. In these models the relation between said stock of knowledge
and age is also obtained. Diamond’s (1984) pioneering article and the
contributions of Levin & Stephan (1991) and Rauber
& Ursprung (2005), among others, can be
highlighted. In general terms, the cited life cycle models reveal the reason
why variables such as the marginal return of research products, the marginal
utility of solving a scientific puzzle, the rates of learning and the
depreciation of knowledge, affect the researcher’s intellectual output and why
it declines with age.
In this section,
a life cycle model for professors in public universities in Colombia is
formulated, assuming that they can divide their time between teaching in the
public university, which is compulsory, research, and other alternative
activities, for example, classes in a private university or consulting, which
incorporates the opportunity cost of the time dedicated to research. The model
is an adaptation in discrete time of the model by Levin and Stephan (1991). A
modification is made because of the fact that future reward does not depend on
the researcher’s prestige, but rather their expectation of future salary depends
on the intellectual productivity accumulated during their whole academic life
cycle.
As
in Levin and Stephan (1991), it is assumed that the profit function of public
university professors in Colombia has as its basis the status of the professor
in the university community,〖Status〗_t, and consumer
goods X_t.
The
The
professor maximizes the current value of their profit during a period of time,
T.
Subject to:
Where
The function of research production is
identified as:
The choice
variables are
Table 1. Values of the parameters for the simulation
of the base line model
|
Parameters |
Value |
|
Profit Function |
|
|
|
0,2 |
|
|
0,8 |
|
Output Function |
|
|
b |
0,8 |
|
Knowledge update |
|
|
a |
0,2 |
|
d |
0,1 |
|
Discount rate |
|
|
|
0,04 |
|
|
0,5 |
|
|
0,1 |
|
Others |
|
|
|
3000 |
|
|
10000 |
|
|
100 |
|
|
2 |
Source: elaborated by the author
Table 2 shows
the first simulation of the model. It is observed that, until the age of 51,
the representative professor of the model dedicates more than one fraction (
Table 2. Simulation of the model –
base line
|
Age |
Status |
X |
Profit |
|
|
|
A |
K |
P |
W |
OW |
|
31 |
1,46 |
6,22 |
2,44 |
1,00 |
2,444 |
0,80 |
0,0 |
2 |
1,46 |
3000,0 |
1978,7 |
|
32 |
1,85 |
6,54 |
2,56 |
0,96 |
2,464 |
0,51 |
2781,9 |
2,238 |
1,12 |
3146,0 |
4871,4 |
|
33 |
2,03 |
6,88 |
2,62 |
0,92 |
2,427 |
0,48 |
5993,4 |
2,349 |
1,10 |
3257,6 |
5183,2 |
|
34 |
2,16 |
7,24 |
2,68 |
0,89 |
2,379 |
0,48 |
9330,8 |
2,445 |
1,14 |
3368,0 |
5165,0 |
|
35 |
2,28 |
7,62 |
2,72 |
0,85 |
2,329 |
0,49 |
12716,2 |
2,544 |
1,20 |
3482,3 |
5068,1 |
|
36 |
2,40 |
8,02 |
2,77 |
0,82 |
2,279 |
0,51 |
16109,2 |
2,649 |
1,26 |
3602,2 |
4937,0 |
|
37 |
2,54 |
8,44 |
2,82 |
0,79 |
2,228 |
0,52 |
19483,4 |
2,764 |
1,34 |
3728,7 |
4789,7 |
|
38 |
2,69 |
8,89 |
2,87 |
0,76 |
2,178 |
0,54 |
22819,7 |
2,889 |
1,42 |
3862,6 |
4636,2 |
|
39 |
2,85 |
9,36 |
2,91 |
0,73 |
2,129 |
0,55 |
26095,6 |
3,026 |
1,51 |
4004,5 |
4475,4 |
|
40 |
3,03 |
9,85 |
2,96 |
0,70 |
2,079 |
0,57 |
29283,5 |
3,176 |
1,61 |
4155,4 |
4304,8 |
|
41 |
3,23 |
10,37 |
3,01 |
0,68 |
2,031 |
0,59 |
32356,1 |
3,340 |
1,71 |
4316,0 |
4127,6 |
|
42 |
3,45 |
10,92 |
3,05 |
0,65 |
1,983 |
0,61 |
35287,9 |
3,520 |
1,83 |
4487,4 |
3948,7 |
|
43 |
3,68 |
11,51 |
3,10 |
0,62 |
1,935 |
0,62 |
38054,0 |
3,718 |
1,96 |
4670,6 |
3771,2 |
|
44 |
3,93 |
12,12 |
3,14 |
0,60 |
1,887 |
0,64 |
40627,8 |
3,933 |
2,09 |
4866,3 |
3596,6 |
|
45 |
4,21 |
12,76 |
3,19 |
0,58 |
1,841 |
0,66 |
42982,1 |
4,168 |
2,24 |
5075,7 |
3427,4 |
|
46 |
4,50 |
13,45 |
3,23 |
0,56 |
1,795 |
0,67 |
45091,1 |
4,423 |
2,39 |
5299,6 |
3268,2 |
|
47 |
4,80 |
14,17 |
3,28 |
0,53 |
1,749 |
0,69 |
46931,3 |
4,699 |
2,56 |
5539,0 |
3124,5 |
|
48 |
5,12 |
14,92 |
3,32 |
0,51 |
1,704 |
0,70 |
48481,5 |
4,996 |
2,72 |
5794,6 |
3002,1 |
|
49 |
5,45 |
15,73 |
3,36 |
0,49 |
1,659 |
0,71 |
49722,0 |
5,313 |
2,89 |
6066,7 |
2905,8 |
|
50 |
5,78 |
16,57 |
3,40 |
0,47 |
1,615 |
0,72 |
50634,0 |
5,648 |
3,06 |
6355,8 |
2839,9 |
|
51 |
6,11 |
17,46 |
3,44 |
0,46 |
1,572 |
0,72 |
51198,9 |
6,001 |
3,22 |
6661,6 |
2808,7 |
|
52 |
6,43 |
18,40 |
3,48 |
0,44 |
1,529 |
0,72 |
51397,9 |
6,367 |
3,37 |
6983,7 |
2816,6 |
|
53 |
6,73 |
19,39 |
3,52 |
0,42 |
1,487 |
0,71 |
51211,3 |
6,743 |
3,51 |
7321,2 |
2867,5 |
|
54 |
6,99 |
20,44 |
3,56 |
0,41 |
1,445 |
0,70 |
50616,0 |
7,123 |
3,63 |
7672,5 |
2963,8 |
|
55 |
7,22 |
21,55 |
3,60 |
0,39 |
1,404 |
0,69 |
49582,9 |
7,500 |
3,72 |
8035,6 |
3106,1 |
|
56 |
7,39 |
22,71 |
3,64 |
0,38 |
1,364 |
0,67 |
48072,1 |
7,866 |
3,78 |
8407,8 |
3292,0 |
|
57 |
7,51 |
23,94 |
3,67 |
0,36 |
1,324 |
0,65 |
46029,8 |
8,215 |
3,81 |
8786,1 |
3517,0 |
|
58 |
7,56 |
25,24 |
3,71 |
0,35 |
1,285 |
0,62 |
43383,6 |
8,537 |
3,81 |
9167,3 |
3774,1 |
|
59 |
7,55 |
26,61 |
3,74 |
0,33 |
1,247 |
0,59 |
40038,9 |
8,825 |
3,77 |
9547,8 |
4055,0 |
|
60 |
7,47 |
28,05 |
3,77 |
0,32 |
1,209 |
0,56 |
35876,2 |
9,072 |
3,70 |
9924,4 |
4350,7 |
|
61 |
7,33 |
29,58 |
3,80 |
0,31 |
1,172 |
0,53 |
30750,8 |
9,274 |
3,60 |
10294,0 |
4654,3 |
|
62 |
7,14 |
31,19 |
3,83 |
0,30 |
1,136 |
0,50 |
24505,3 |
9,426 |
3,47 |
10653,9 |
4974,1 |
|
63 |
6,84 |
32,88 |
3,86 |
0,29 |
1,100 |
0,46 |
17037,6 |
9,525 |
3,27 |
11001,0 |
5387,6 |
|
64 |
6,20 |
34,67 |
3,88 |
0,27 |
1,064 |
0,38 |
8563,2 |
9,552 |
2,79 |
11327,8 |
6233,3 |
|
65 |
4,60 |
36,59 |
3,87 |
0,26 |
1,021 |
0,18 |
0,0 |
9,433 |
1,50 |
11606,3 |
8244,2 |
Source: elaborated by the author.
Figure 1 shows
the dynamics of the stock of knowledge and the number of publications during
the life cycle of the research professor. It is observed that although the
professor accumulates knowledge during almost their whole life as a university
professor, their academic production declines as from 58 years of age. This
happens because, as it can be seen in the trajectory
Figure 1. Dynamics of the stock of knowledge
and publications
Source: elaborated by the author
If the starting
salary increases to
If added to the
increase in the initial salary, the value of the salary point increases too,
and VSP goes from 100 to 110, the total output of a researcher during their
life cycle rises to
Figure 2 shows
the graphs of the time fractions dedicated to research on the base line, and when
there are exogenous changes in the salary and in the value of salary point. It
is observed that as the salary and the value of the salary point increase
exogenously, the fraction of time dedicated to research is higher for every
year. In the two first cases, the highest peak of the fraction of time
dedicated to research is reached at 51 years of age. The professor dedicates
all their available time to research until they are 55 when the value of the
salary point increases.
Figure 2. Time fraction dedicated to
research in each simulation
Source: elaborated by the author
DATA
The aim of this
section is to make a quick description of the data available for the estimation
of the econometric model of the intellectual productivity of professors in
public universities. The data corresponds to the information from the body of
professors of the Universidad del Valle, for the
period of 2004-2012. The information
includes: starting date at the university (year of recruitment), age, salary, faculty
they belong to, academic level, category in the pay scale, research group they
belong to, points for intellectual output, points for outstanding teaching
practices, and points for academic-administrative managerial activities of the
professors at the Universidad del Valle, subject to Decree 1279, which
establishes the pay and tax scale of professors in state universities in
Colombia.
Table 3 shows
the average points for academic productivity per faculty, resulting from the
annual salary update and its standard deviation. The highest average
corresponds to the professors of the science faculty and the lowest is that of
the school of psychology.
Table 3. Annual average of the points for productivity per
faculty
|
Faculty |
Obs. |
Annual Average |
Statistical Deviation |
Binary Variation Representation |
|
Arts |
806 |
3,09 |
9,31 |
|
|
Administration |
302 |
4,23 |
13,04 |
|
|
Engineering |
1053 |
6,64 |
16,24 |
|
|
Socio-economics |
176 |
5,65 |
12,29 |
|
|
Sciences |
753 |
10,18 |
21,85 |
|
|
Psychology |
145 |
2,91 |
8,63 |
|
|
Health |
1158 |
3,44 |
11,22 |
|
|
Humanities |
781 |
3,88 |
8,54 |
|
Source: elaborated by the author
Table 4 presents
the average points for academic productivity per academic level and its
standard variation. The professors with a doctorate show the highest
productivity, followed by professors who hold a master’s degree (as was
expected), that of the former being more than two and a half times higher than
the latter. Professors with a specialization do not appear to have a research
advantage over professors who hold a bachelor’s degree, as their degree of
productivity is lower.
Table 4. Average annual points for
academic productivity (per) academic level
|
Academic Level |
Obs. |
Annual
average |
Statistic
deviation |
Binary Variation Representation |
|
Undergraduate |
650 |
2,89 |
9,60 |
|
|
Specialization |
925 |
1,98 |
6,86 |
|
|
Master’s Degree |
2101 |
3,81 |
10,27 |
|
|
Doctorate |
1498 |
10,10 |
20,83 |
|
Source: elaborated by the author
Table 5 presents
the average points for academic productivity, according to the category of the
professor and their standard deviation. It is observed that the average points
for academic productivity of the professors’ increases in accordance with the
academic degree reached.
Table 5. Average annual points for intellectual
productivity according to category
|
Category |
Obs. |
Annual
Average |
Statistical
deviation |
Representation Binary Variation |
|
Auxiliary |
921 |
2,20 |
5,85 |
|
|
Assistant |
1629 |
4,21 |
12,00 |
|
|
Associate |
909 |
5,65 |
13,09 |
|
|
Tenure |
1715 |
7,49 |
18,50 |
|
Source: elaborated by the author
Table 6 displays
the descriptive statistics of the average of the age of the professors, the
real salary at 2007 prices, and the points accumulated for outstanding teaching
practices and academic productivity during the years 2004-2011.
Table 6. Descriptive statistics of the
continuous variables
|
Variable |
Obs. |
Annual Average |
Statistic deviation |
Min. |
Max. |
Representation |
|
Age |
6059 |
48,7 |
8,6 |
25 |
84 |
|
|
Real salary |
6059 |
3’526.821 |
1’763.153 |
824.942 |
15’100.000 |
|
|
Points for outstanding teaching practices |
6059 |
7,9 |
8,7 |
0,0 |
50,0 |
|
|
Points for intellectual
productivity |
6059 |
91,5 |
135,9 |
0,0 |
1350,9 |
|
Source: elaborated by the author
ESTIMATION OF THE ECONOMETRIC MODEL AND ANALYSIS OF
THE RESULTS
The
specification of the econometric model for intellectual output is the
following:
Where:
The vector
The expression
for the calculation of the
In (9), the term
VSP corresponds to the value of the salary point, NAP is the number of annual
payments that the professor receives, and
The
specification of the model (8) proposes a relation between real salaries and
intellectual productivity which has the shape of an inverted U and follows Tang
(2000). The reason that the author suggests is the following: to begin with,
the initial increase in the salaries of the employees makes them more
productive, because said increase makes leisure more expensive. Nevertheless,
the additional increase in the salaries will reduce productivity, because the
employees get richer and allow themselves more free
time.
So as to have a
first approximation of the factors that determine professors’ publications, a Logit model is estimated where the dependent variable is
the binary variable
The estimation of the Logit model appears in
Table 7. The age, salary, points for intellectual productivity and outstanding
teaching practices are left aside, in order to take into account the moment in
which the decision was made to do research with the aim of publishing the
article or book of interest. The variables of the professors’ pay scale
category, the real salary lagged in levels, and the binary variables of the
faculties are not significant, except for the humanities and psychology
dummies.
Table 7. Logit model of random
effects to establish the determinants of publication
|
|
|
|||
|
Observations |
2509 |
|||
|
Number
of groups |
771 |
|||
|
|
|
|||
|
Explanatory variable |
Coefficient |
Standard
error |
|
|
|
|
0.1492 |
0.1888 |
0.79 |
0.430 |
|
|
0,4376 |
0.2570 |
1.70 |
0.089 |
|
|
-0.0582 |
0.3229 |
-0.18 |
0.857 |
|
|
0.4956 |
0.2340 |
2.12 |
0.034 |
|
|
0.6836 |
0.2122 |
3.24 |
0.001 |
|
|
1.3992 |
0.2467 |
5.67 |
0.000 |
|
|
0.8850 |
0.1417 |
6.24 |
0.000 |
|
|
-0.050 |
0.0095 |
-5.27 |
0.000 |
|
|
-5.07e-09 |
1.67e-07 |
-0.03 |
0.976 |
|
|
-6.63e-14 |
1.82e-14 |
-3.64 |
0.000 |
|
|
0.0126 |
0.0016 |
7.85 |
0.000 |
|
|
0.0216 |
0.0088 |
2.45 |
0.014 |
|
|
0.7764 |
0.2162 |
3.59 |
0.000 |
|
|
-0.7529 |
0.4008 |
-1.88 |
0.060 |
|
|
-0.0140 |
0.2114 |
-0.07 |
0.947 |
|
|
0.2115 |
0.3480 |
061 |
0.543 |
|
|
-0.1143 |
0.2930 |
-0.39 |
0.696 |
|
|
0.3244 |
0.2113 |
1.53 |
0.125 |
|
|
0.0287 |
0.2476 |
0.12 |
0.908 |
|
|
0.3023 |
0.1503 |
2.01 |
0.044 |
|
|
0.2590 |
0.1562 |
1.66 |
0.097 |
|
|
0.5817 |
0.1710 |
3.40 |
0.001 |
Source: elaborated by the author
Table 8
presents, in columns (II) and (III), the estimation of the life cycle and intellectual
output models, through a Tobit estimation of random effects. The only
difference between the models is that the first uses age as a regressor, and the second uses the incentive for
intellectual output as an explanatory variable of interest.
On the other hand,
there can be some bias in the estimators of the Tobit model of random effects
if the non-observed individual heterogeneity has the shape of fix/steady
effects correlated with the explanatory variables. In order to overcome this
problem, Levin and Stephan (1991) estimate a Tobit model which includes
individual dummies, Goodwin & Sauer (1995) and Rauber
& Ursprung (2008) incorporate a categorical
variable, the result of ranking researchers in quintiles from the average
production of their life cycle. Following a similar strategy, Vella and Verbeck (1997) use the
residuals of a preliminary regression to rank researchers based on their
research ability, not captured by the deterministic part of the model. In
columns (IV) and (V), the estimations of the life cycle and intellectual output
models appear, using the Vella and Verbeck (1997) methodology to capture the fixed individual
heterogeneity.
The discount
rates used in the models of intellectual output were:
As was to be
expected, the more the academic level of the professor increases, the more
their intellectual output improves. The estimations show that a professor with
a doctorate has approximately 6 points more for intellectual output than a
professor who holds a master’s degree, and almost 8
points more than a professor with a specialization.
Being part of a
research group has a positive effect on intellectual output of approximately 6
points a year.
The effect of age
is negative, because the older the professor gets, the profit expected from
publications drops. Even when with the passing of time the research professor
improves their stock of knowledge and, therefore, their publishing abilities,
this positive effect is captured by the following variables: lagging accumulated intellectual
productivity (
The effect of
the incentive is positive and significant. The idea of including the variables
(
(
It is also
important to highlight that, despite the incorporation of the average
production of a professor during their life cycle (residual) as an explanatory
variable of the Tobit model, the variable (
The Faculty of
Science features ceteris paribus, the
greatest intellectual output, which would be nearly 3 points above the average
of the university. In the Tobit estimation of random effects, the lowest ceteris paribus intellectual output is
found in the Institute of Psychology. In Vella and Verbeck’s (1997) Tobit
estimation, the School of Science produces the lowest ceteris paribus intellectual output.
The variable
lagging real salary in levels is not significant in any of the estimations, and
the same variable squared turned out to be significant (only in the Tobit
estimation of random effects) as well as its negative effect, because when a
professor receives a high salary, they prefer leisure over income or status in
the university.
Binary period
variables are included to capture the cohort effect following the
recommendation of Rauber and Ursprung
(2005). The time dummies reveal the rapid growth in the intellectual output of the
Universidad del Valle professors, not associated with
the individual characteristics observed.
Table 8. Estimation of life cycle
and intellectual output models
|
Dependent variable |
Points
for intellectual output ( |
|||
|
Number of observations |
2509 |
|||
|
Explanatory variable |
Tobit
estimation of random effects |
Tobit
estimation – Vella and Verbeck methodology (1997) |
||
|
Life cycle
Model (II) |
Intellectual output (III) |
Life cycle
Model (IV) |
Intellectual output (V) |
|
|
|
3.1452 |
3.1230 |
3.0889 |
3.0739 |
|
0.077 |
0.088 |
0.049 |
0.050 |
|
|
|
6.1391 |
6.1207 |
4.9007 |
4.9032 |
|
0.000 |
0.000 |
0.000 |
0.000 |
|
|
|
12.6735 |
12.6659 |
10.9312 |
10.9147 |
|
0.000 |
0.000 |
0.000 |
0.000 |
|
|
|
6.1416 |
6.1376 |
5.8459 |
5.8402 |
|
0.000 |
0.000 |
0.000 |
0.000 |
|
|
|
-0.5108 |
- |
-0.5333 |
- |
|
0.000 |
0.000 |
|||
|
|
- |
3.16e-6 |
- |
3.44e-6 |
|
0.000 |
0.000 |
|||
|
|
-6.03e-13 |
-6.04e-13 |
- |
- |
|
0.000 |
0.000 |
|||
|
|
0.1285 |
0.1286 |
0.0576 |
0.0576 |
|
0.000 |
0.000 |
0.000 |
0.000 |
|
|
|
0.1475 |
0.1475 |
0.0962 |
0.0959 |
|
0.019 |
0.020 |
0.071 |
0.072 |
|
|
|
- |
- |
1.1096 |
1.1096 |
|
0.000 |
0.000 |
|||
|
|
3.261 |
3.2567 |
2.7849 |
2.7754 |
|
0.011 |
0.012 |
0.006 |
0.007 |
|
|
|
- |
- |
-2.5572 |
-2.5540 |
|
0.027 |
0.027 |
|||
|
|
-5.6811 |
-5.6844 |
- |
- |
|
0.060 |
0.060 |
|||
|
|
2.9069 |
2.9276 |
2.0193 |
2.0251 |
|
0.013 |
0.014 |
0.072 |
0.071 |
|
|
|
3.9471 |
3.9617 |
3.3067 |
3.3147 |
|
0.001 |
0.001 |
0.004 |
0.003 |
|
|
|
8.1645 |
8.1932 |
6.3004 |
6.3135 |
|
0.000 |
0.000 |
0.000 |
0.000 |
|
Source: elaborated by the author
The estimated
effects of the incentive for intellectual output were
for the Tobit model
of random effects and
Figures 3 and 4
show the estimated effect of the incentive ceteris
paribus, over the intellectual productivity of the professors of the Universidad
del Valle, calculated from the random effects Tobit
model (Figure 3) and the Vella and Verbeck Tobit model (Figure 4).
In the first, the effect is 20.3 points for a 30-year-old professor, 15.27
points for a 40-year-old, 10.18 for a 50-year-old, and
5 points for a 60-year-old. In the second the effect is 15.4 points for a
30-year-old professor, 12.59 points for a 40-year-old, 9.15 points for a 50-year-old, and almost 5 points for a 60-year-old
professor.
Figure 3. Estimated effect of the incentive over intellectual output calculated
based on the random effects Tobit estimation.
Source: elaborated by the author
Figure 4. Estimated effect of the incentive over intellectual output calculated
based on the Vella and Verbeck
Tobit estimation (1997).
Source: elaborated by the author
Figure 5 reveals
the estimated effects using the results of the Tobit model with fix effects of
increments of 10 %, 20 %, 30 %, 40 % and 50 % in the value of the salary point
over intellectual output, associated with each age level.
effect effect effect effect
Figure 5.
Estimated effect of the percentage increase of the salary point value
Source: elaborated by the author
CONCLUSIONS
In the year
2012, out of 759 professors covered by Decree 1279, 23.8% were professors
without a Master’s degree or a doctorate, 44.3% were professors with Master’s
degree, and 31.8% had a doctorate. The estimations found that a professor with
a doctorate has approximately 6 more points for intellectual output than a
professor who only holds a Master’s degree, and almost 8 more points than a
professor who only has a specialization. The first difference is equivalent to
the publication of a couple of C type articles a year and the second to one B
type article a year. Therefore, there exists the possibility to increase intellectual
output, through the policy of study commissions to improve the academic level
of professors.
It was discovered that participation in a
research group affects intellectual output by almost 6 points, a similar
contribution to the difference in intellectual output between a professor with
a doctorate and one with a Master’s degree. For this reason, it is important
that the authorities foster professorial participation in research groups.
In the
theoretical model developed in this article, the salary of a professor affects
their intellectual output, because the higher their salary in the public
university, the less time the professor dedicates to other activities that
generate income, such as, for example, working at another university or consulting.
Nevertheless, the empirical evidence obtained from the estimations reveals that
the level of salary does not have a positive effect on the intellectual output
of professors in public universities. This can be considered an indicator that
there is no tradeoff between the time dedicated to research and that dedicated
to other activities that generate income.
In the article
of Levin and Stephan (1991), the salary of a professor turned out to be
significant in the explanation of their research productivity, but it has to be
highlighted that, in that estimation, salary was used as a proxy of the future
profits of the professor’s publications. In this article, there is an
approximate measurement of the marginal utility of the research (the current
value of the salary point).
However, given
that the estimation sample covers only professors from the Universidad del
Valle, a public university with relatively low starting salaries, the effect of
the salary in hiring teachers with better research abilities cannot be
determined and, therefore, it's effect on intellectual output cannot be
determined either. Due to this limitation, what can be observed is that the salary
level does not affect intellectual output through the distribution of time
dedicated to research.
The estimations
found an average effect of the salary incentive over productivity of 15.4
points for a 30-year-old professor, 12.59 points for a 40-year-old professor,
9.15 points for a 60-year-old professor, and almost 5 points for a 60-year-old professor,
etc. What is clearly revealed is the importance of the incentive factor, if we
compare it with the difference estimated for the effect on intellectual output,
between having a doctorate and a master's degree which provides around 6
points, or being part of a research group which provides 6 points.
The best
estimation in the Tobit model of Vella
and Verbeck (1997) was obtained with a 1.9% discount
rate, and in the Tobit estimation with random effects without discount rate.
Given the low magnitude of the discount rate, it can be concluded that the
strategy of some private universities of offering premiums for intellectual
output over a limited period (a year) does not compete with the incentive of
the public university, because a very high amount of the marginal value of the
premium for intellectual output would be required, in order to balance the time
during which the professors of public universities receive their compensation,
especially in the case of young professors.
One of the main
problems that public universities in Colombia face when hiring new professors
is the legal restrictions in setting a starting salary that can attract the
most talented candidates in the market[4]. However,
teaching in a Colombian public university also has a couple of aspects in its favor:
i. Exclusive dedication is not required, ii. Depending on labor seniority,
experience, and, above all, their publications, a professor in a public
university may get to earn the same as or more than a professor in a private
university.
If the
advantages that a public university offers compensate the low starting salary
of the professors of public universities, the distribution of research abilities
of the professors hired would not be affected, and in that case, it could be
affirmed that the salary incentive policy is preferred to that of high
salaries.
In this sense,
policies such as the position of full time professor may positively affect
research, because it may help to reduce the starting salary restriction which
enables the hiring of professors with greater research abilities, as it allows
the salary of the professor to increase by up to 22%. However, a direct effect
of the highest salary on effort in research and intellectual output would not
be expected.
The significance
and the effect of the time dummies
also showed the ceteris paribus growth
of the intellectual output of the professors in the Universidad de Valle,
probably due to cohort effects and a greater opportunity to publish, associated
with a better supply of specialized journals or journals better ranked by Colciencias, or both.
REFERENCES
Aksnes, D.,
Rorstad, K., Piro, F. & Sivertsen, G. (2011). Age and
scientific performance. A large-scale study of Norwegian
scientists. In Paper presented at the ISSI 2011,
Durban, South Africa.
Aksnes, D. W., Rorstad, K., Piro, F. & Sivertsen, G. (2011b). Are female researchers less cited? A largescale study of Norwegian scientists. Journal of the American Society for
Information Science and Technology, 62(4),
628–636.
Allison, P. & Stewart, J. (1974).
Productivity differences among scientists: evidence for accumulative
advantage. American Sociological
Review, 39, 596–606.
Archibugi, D. & Coco,
A. (2004). A new indicator of technological capabilities for
developed and developing countries. World
Development, 32(4), 629–654.
Astin, H. &
Bayer, A. (1979). Pervasive sex differences in the academic reward
system: Scholarship, marriage, and what else? In D.R. Lewis & W.E. Becker
(Eds.), Academic rewards in higher
education. Cambridge, MA: Ballinger.
Baccini, A., Barabesi, L., Cioni, M. & Pisani, C. (2014). Crossing the
Hurdle: The determinants of individual scientific performance. Scientometrics,
1-28.
Bayer, A. & Dutton, J. (1977).
Career age and research-professional activities of academic scientists: tests
of alternative nonlinear models and some implications for higher education
faculty policies. The Journal of Higher Education, 48(3), 259-282.
Bonaccorsi,
A. & Daraio, C. (2003).
Age effects in scientific productivity - the case of the Italian national
research council (CNR). Scientometrics, 58(1), 49-90.
Carayol,
N. & Matt, M. (2006). Individual
and collective determinants of academic scientists on productivity. Information Economics and Policy, 18, 55–72.
Carvajalino, A. & Ariza, A. (2008). Evaluación
de los incentivos a la investigación y extensión en la Universidad Industrial
de Santander. Tesis de grado. Facultad de Ciencias Humanas. Escuela de
Economía y Administración. Universidad Industrial de Santander.
Cohen, W., Florida,
R. & Goe, W. (1994). University-industry
research centres in the United States. Pittsburgh, PA:
Centre for Economic Development, Carnegie Mellon University.
Cole,
S. (1979). Age and scientific performance. American
Journal of Sociology, 84(4),
958-977.
Cole, J. &
Cole, S. (1973). Social
stratification in science. Chicago: Chicago University Press.
Cole, S., Cole, J. &
Dietrich, L. (1978). Measuring
the Cognitive State of Scientific Disciplines. In
Y. Elkana, J. Lederberg, R. Merton, A. Thackray & H. Zuckermann. (Eds.).
Towards a metric of science: the advent of science indicators, (pp.
209-251). New York: Wiley.
Cole,
J. & Zuckerman, H. (1984). The productivity puzzle: Persistence and change in
patterns of publication of men and women scientists. In M. W.
Steimkamp & M. Maehr
(Eds.). Advances in motivation and achievement, 2, 217–258.
Crane, D.
(1965). Scientists at major and minor universities: a study of productivity and
recognition. American Sociological
Review, 30(5), 699-714.
Creswell, J. W.
(1986). Measuring
faculty research performance. New Directions for Institutional Research 50. San Francisco: Jossey-Bass.
David, P.
(1994). Positive feedbacks and research productivity in science: reopening
another black box. In O. Grandstrand
(Ed), Economics and Technology.
Amsterdam: Elsevier.
Dundar,
H. & Lewis, D. (1998). Determinants
of research productivity in higher education. Research in Higher
Education, 39(6), 607-631.
Fielden, J. &
Gibbons, J. (1991). Merit myopia and business school faculty publications. Business Horizons, 34(2), 8-12.
Fox, M. (1992).
Research, teaching, and publication productivity: mutuality versus competition
in academia. Sociology of Education, 65(4), 293-305.
Gonzalez, C. & Veloso,
F. (2007). The determinants of research output and impact: A
study of Mexican researchers. Research Policy, 36, 1035-1051.
Goodwin,
T. & Sauer, R. (1995). Life cycle productivity in academics research:
Evidence from cumulative publication histories of academic economists. Southern Economic Journal, 61(3), 728-743.
Gorbaneff, Y. (2003). Teoría del agente-principal y el
mercadeo. Revista EAFIT, 129, 75-86.
Gutiérrez, J. & Berrío,
O. (2011). Punto de inflexión entre empresas y universidades ante la relación
universidad, empresa y Estado en Colombia. Revista
Universidad & Empresa, 21,
167-191.
Guzmán, A. & Trujillo, M.
(2011). Políticas de incentivos relacionadas con la investigación: una revisión
crítica desde la teoría de contratos. Estudios
Gerenciales, 27(120),
127-145.
Jordan,
J. M., Meador, M. & Walters, S. (1989). Academic research productivity,
department size and organization: further results. Economics of Education Review, 8(4),
345-352.
Knorr, K.,
Mittermeier, R., Aichholzer, G. & Waller, G. (1979). Individual
publication productivity as a social position effect in academic and industrial
research units. In F. Andrews (Ed.), The effectiveness of research groups in six countries (pp.
55–94). Cambridge: Cambridge University
Press.
Kyvik, S. (1991). Productivity in Academia. Scientific publishing at
Norwegian universities. Oslo: Universitetsforlaget.
Kyvik,
S. (1995). Are big university departments better than small
ones? Higher Education, 30(3),
295-304.
Kyvik, S. & Teigen, M. (1996). Child care, research
collaboration, and gender differences in scientific productivity. Science, Technology & Human Values, 21(1), 54–71.
Lehman,
H. C. (1958). The chemist most creative years. Science, 127, 1213–1222.
Lehman,
H. C. (1960). The decrement in scientific productivity.
American Psychologist, 15, 128–134.
Levin, S. & Stephan, P. (1989).
Age and research productivity of academic scientists. Research in Higher Education, 30(5), 531–549
Levin,
S. & Stephan, P. (1991) Research productivity over the life cycle: evidence
for academic scientists. The American Economic Review, 81, 114–132.
Long, J. S.
(1978). Productivity and academic positions in the scientific
career. American Sociological
Review, 43, 889–908.
Lotka A. (1926). The frequency distribution of scientific productivity. Journal of the Washington Academy of Sciences, 16, 317-323.
Manjarrés, L. (2009). Las relaciones
universidad empresa y su efecto sobre la segunda misión universitaria. Tesis doctoral, Departamento de estadística e
investigación operativa aplicada y calidad. Universidad Politécnica de Valencia
Merton, R.K.
(1968). The Matthew effect in science. Science, 159(3810), 56-63.
Miller, A. & Serzan,
S. (1984). Criterion for identifying a refereed journal.
Journal of Higher
Education, 55(6), 763-699.
Obembe, O. (2012). Determinants of scientific productivity among Nigerian University
academics. Indian
Journal of Science and Technology, 5(2),
2155-2164.
Paasi, A. (2005). Globalization, academic capitalism, and the uneven geographies of
international journal publishing spaces. Environment and Planning,
37, 769–789.
Perozo, S., De Arteaga, F. & Fuenmayor,
B. (2008). La productividad investigativa de los docentes del Instituto
Universitario de Tecnología de Cabimas. Revista
NEGOTIUM, 3(9), 72-87.
Piedra, Y. & Martínez, A. (2007). Producción
científica. Revista Ciencias de la
información, 38(3), 33-38.
Porter,
S. & Umbach, P. (2001). Analyzing faculty workload data using multilevel modeling. Research in Higher Education, 42(2), 171-196.
Print,
M. & Hattie, J. (1997). Measuring Quality in Universities:
An Approach to Weighting Research Productivity. Higher Education, 33(4),
453-469.
Rauber,
M. & Ursprung, H. (2008).
Life cycle and cohort productivity in economic research: the case of Germany. German
Economic Review, 9(4), 431-456.
Rebne, D. (1990). Determinants of individual productivity: a
study of academic researchers. California: Institute
of Industrial Relations Publications
Center, UCLA.
República
de Colombia (1992). Decreto 1444 de
septiembre 3 de 1992.
Por el cual se dictan disposiciones en materia salarial y prestacional para los
empleados públicos docentes de las universidades públicas del orden nacional.
República de Colombia (2002). Decreto 1279 de Junio 19 de 2002. Por el
cual se establece el régimen salarial y prestacional de los docentes de las
Universidades Estatales.
Reskin, B. (1978). Scientific productivity, sex, and location in the institution of
science. American Sociological
Review, 83(5), 1235-1243.
Smeby, J. & Try,
S. (2005). Departmental contexts and faculty research activity in Norway. Research in Higher Education, 46(6), 593-619.
Stephan, P.
(2010). The economics of science. In
Handbook of the economics of innovation. Vol-1, (pp 217-274).
Burlington: Bronwyn H. & Nathan Rosenberg, Acad. Press. .
Tang, C. (2010).
A note on the nonlinear
wages-productivity nexus for Malaysia. Munich, Germany: MPRA Paper,
University Library of Munich.
Weiss,
Y. & Lillard, L. (1982). Output
variability, academic labor contracts, and waiting times for promotion. Research in Labor Economics, 5, 157–188.
Zuckerman,
H. & Merton, R. (1972). Age, aging, and age structure in
science. In M. Riley, M. Johnson & A. Foner
(Eds.), A sociology of age
stratification: Aging & society (Vol. 3). New York: Russel Sage
foundation.
Zuckerman,
H. & Merton, R. (1973). The Sociology
of Science: Collected Papers of R.K. Merton, (Reimp.). Chicago, USA: Chicago University, Chicago Press.
* This
article is part of the thesis “Salaries
and Work Productivity in Colombia” to opt for the title of the Doctorate in
Development Economics of the Latin American Faculty of Social Sciences, Ecuador
branch.
**
Assistant professor, Economics Department, Universidad del Valle, Cali,
Colombia. Civil Engineer from the Universidad Francisco de Paula Santander.
Master’s Degree in Economics from the Universidad Javeriana.
Master’s Degree in Environmental Economics and Natural
Resources from the Universidad de los Andes. Jhon.mendez@correounivalle.edu.co
***
Associate Professor. School of Economics, Universidad Central
de Venezuela, Caracas, Venezuela. Ph.D. University of London,
England. leoverave@gmail.com
[1]
In the year 2014, the average starting salary of a professor who holds a
doctorate was only 3.8 million Colombian pesos on average.
[2]
In the simulation, it is assumed that each publication corresponds to a salary point, in fact each publication is valued on the basis of
article 10 of Decree 1279.
[3]
Each period corresponds to two years, assuming cycles of research-publication
every couple of years.
[4]In
the year 2014, the average starting salary of a professor with a doctorate was
only 3.8 million pesos a year, on average.