The Investigation of Artificial Intelligence Application in the Public Administration s Literature

Transcription

1 The Investigation of Artificial Intelligence Application in the Public Administration s Literature Srirath Gohwong Faculty of Social Sciences, Kasetsart University, Thailand srirathg3@yahoo.com Abstract This article is the study of Artificial Intelligence (AI) application for Public Administration (PA) by investigating many key writings of PA scholars in six paradigm s Henry and three Minnowbrook s conferences. The findings are that AI is not paid attention to PA scholars as a tool for IT-based reform such as big data management and knowledge management for public policy analysis. Keywords: Artificial Intelligence, Knowledge Management, Minnowbrook Conferences, Public Administration, Public Policy Analysis, Introduction Artificial Intelligence (AI) is the field of research that studies and develops of intelligent machines and software for replacing human thinking and deciding. (Simon, 1977; Dervojeda et.al., 2013). It play a big role in e-government in the age of chaos, especially big data-driven decision-making and knowledge management (KM) for public policy analysis. (Yick, 2009; O Leary, Van Slyke and Kim, 2010; Laudon and Laudon, 2014) However, there is no study about the development of AI s application in the writings of Public Administration (PA). Hence, the one and only one objective of this article is to investigate the development of AI application in the history of PA. The Paradigm Shift of PA Paradigm is one of the most important words for PA because the easiest way to see the development of the field by classifying theories of PA into eras. Though it is employed by many scholars such as Henderson (1966), Waldo (1968), Ostrom (1974), Henry (1975, 2010), Golembiewski (1977), Bozeman (1978), McCurdy (1986) (Bowornwathana, 1984; Shafritz and Hyde, 1997; Henry, 1975, 2010). However, this article will employ only Henry s work, which is the one of the Public Administration Review (PAR) s 75 Most Influential Articles (PAR, 2015). The six paradigm of Henry are as follows: Paradigm I-Politics-Administration Dichotomy ( ), Paradigm II-Principle of Administration ( ), The Challenge ( ), The Reaction to the Challenge ( ), Paradigm III-Public Administration as political science ( ), Paradigm IV-Public Administration as Management science ( ), Paradigm V-Public Administration as Public Administration/ Search for curricular Autonomy (1970 onwards), and Paradigm VI-Governance (1990-onwards). (Henry, 1975, 2010) AI in the Six Paradigm of Henry AI is simply defined as the application of computer hardware and software for making a nonprogrammed decision, which applies specific solutions for an unstructured problem. This [8]

2 kind of problem is new, unique, nonrecurring with information deficiency. (Schermerhorn, Jr., 2013; Robbins, DeCenzo and Coulter, 2013) With the investigation of the development of AI application in PA by using six paradigm of Henry, I categorize the era of AI into two eras as follows: pre-ai oriented PA era and AI oriented PA era. Pre-AI Oriented PA Era This era comprises three periods-pre-paradigm period, paradigm I, and paradigm II. The same contribution to AI of these periods is to focus the importance of decision-making for PA. Pre-Paradigm Period Before the paradigm I, the identification of the importance of non-programmed decision, which is the intellectual root of AI, appeared in the first work of PA in 1887, Woodrow Woodrow s The Study of Administration, as the reason of emergence of new field of study. It looks like a candle in the dark because PA scholars have to wait for 58 years for indication of artificiality in decision-making with the well-known book of H.A. Simon in Wilson as the father of PA (1887: ) stated that And those views are steadily widening to new conceptions of state duty; so that, at the same time that the functions of government are everyday becoming more complex and difficult, they are also vastly multiplying in number. Administration is everywhere putting its hands to new undertakings. The utility, cheapness, and success of the government s postal service, for instance, point towards the early establishment of governmental control of the telegraph system. Or, even if our government is not to follow the lead of the governments of Europe in buying or building both telegraph and railroad lines, no one can doubt that in some way it must make itself master of masterful corporations. The creation of national commissioners of railroads, in addition to the older state commissions, involves a very important and delicate extension of administrative functions. Whatever hold of authority state or federal governments are to take upon corporations, there must follow cares and responsibilities which will require not a little wisdom, knowledge, and experience. Such things must be studied in order to be well done. And these, as I have said, are only a few of the doors which are being opened to offices of government. The idea of the state and the consequent ideal of its duty are undergoing noteworthy change; and the idea of the state is the conscience of administration. Seeing every day new things which the state ought to do, the next thing is to see clearly how it ought to do them. This is why there should be a science of administration which shall seek to straighten the paths of government, to make its business less unbusinesslike, to strengthen and purify its organization, and to crown its duties with dutifulness. This is one reason why there is such a science. Paradigm I In this paradigm, there are two scholars with different points of view on decision-making: Goodnow and Weber. Goodnow favored non-programmed decision-making whereas Weber trusted in programmed decision-making. Non-programmed decision-making appeared again in decision of policy of the classical work of Goodnow, Politics and administration: a study in government, in his most famous simple theory of politics and administration dichotomy. According to Goodnow, politics (policy) must be separated from administration. Public policy partly relates to non-programmed decision-making because policy-makers must make a decision under the limitation of incomplete information. He argued that 'The will' of the state or sovereign must be made up and formulated before political action can be had. The will of the state or sovereign must be executed, after it has been formulated, if that will is to [9]

3 result in governmental action. All the actions of the state or its organs, further, are undertaken with the object, either of facilitating the expression of this will or of aiding in its execution. This would seem to be the case whatever may be the formal character, of the governmental system. (Goodnow, 1900: 9) He also stated the incomplete information that No control which a political body can have over a body entrusted with the acquisition of facts and the gathering of information can result in the gathering of more facts or the acquisition of more exact information. (Goodnow, 1900: 80-81) On contrary, Max Weber, a German scholar in the same paradigm with Goodnow, strongly pointed that programmed decision-making for structured program is very important for modern organization in his two famous publications, Theory of Social and Economic Organization, published in 1911 (McCurdy, 1977; Weber, 1969), and Politics as Vocation, published in 1919 (Weber, 1946). Although Weber strongly agree with Goodnow s politics/administration dichotomy by stating that According to his proper vocation, the genuine official and this is decisive for the evaluation of our former regime will not engage in politics. Rather, he should engage in impartial 'administration.' This also holds for the so-called 'political' administrator, at least officially, in so far as the raison d'etat, that is, the vital interests of the ruling order, are not in question. Sine ira et studio, 'without scorn and bias, he shall administer his office. Hence, he shall not do precisely what the politician, the leader as well as his following, must always and necessarily do, namely, fight. (Weber, 1946b: 95), he strongly believed that rule-based management as programmed decision is the solution for modern organization because rules will lead to the consistency of decisionmaking. Weber stated that (4) The rules which regulate the conduct of an office may be technical rules or norms. (Weber, 1969) Paradigm II Decision-making in this paradigm is only programmed decision-making in the works of Henry Fayol, James D. Mooney and Alan C. Reiley, Lyndall Urwick. Urwick with his work in 1934, The Function of Administration with special reference to the work of Henri Fayol, reconciled his idea with the Fayol s the function of administration, and the principles and administrative duties, and Mooney and Riley s logical arrangement of the principles of organization and proposed the analytical framework for the function of administration. Programmed decision-making is one of that function. Urwick stated in 1934 that Fayol s analysis of organization starts with a list of the administrative duties which every organization shall be designed to accomplish. This list is shown in the first column of Table IV Administrative duties.1. See that the plan of operations is carefully prepared and strictly carried out. 5. Make decisions which are clear, distinct and precise. (Gulick and Urwick, 1937: ) AI Oriented Public Administration Era This era can be categorized into two periods as follows: pre-minnowbrook and Minnowbrook. The pre-minnowbrook is the period from which cover the era of the challenge and the reaction of PA ( ), the third paradigm and fourth paradigm. During This era cover the duration of the challenge ( ) and the reaction to the challenge ( ). The 1935 is the beginning of this era because it is the year that Simon was interested in decision-making (Simon and Newell, 1972). In 1949, non-programmed decision-making also appeared in the third edition of Public Administration s Leonard D. White, who wrote the first textbook of Public Administration in 1924 in paradigm I. White stated that The performance of routine duties is not only a good training for the higher posts of the service, but, as a rule, it is a bad training A situation calling for a decision may have initial consideration either at the top or at the bottom of the hierarchy. However, these [10]

4 above-mentioned scholars since 1887 just argued the importance of this type of decisionmaking. The non-programmed decision-making and AI were formally studied by Herbert Alexander Simon. What Simon studied since 1935 was first called decision-making, later called information processing, and finally called computer science (Simon and Newell, 1972). Simon formally indicated the non-programmed decision-making and its connection with AI in his most famous books of all time for PA scholars and PA students in 1945, Administrative Behavior: A Study of Decision-making Processes in Administrative Organization (Simon, 1996). This scholar was 29 years old when his book was published in 1945, 60 years since Wilson s work. Simon (1945: 252; 1996: xii) pointed the problem of artificiality in his Administrative Behavior as follows:... administration is not unlike play-acting. The task of the good actor is to know and play his role, although different roles may differ greatly in content. The effectiveness of the performance will depend on the effectiveness of the play and the effectiveness with which it is played. The effectiveness of the administrative process will vary with the effectiveness of the organization and the effectiveness with which its members play their parts. From the above mentioned page of Simon, he believed that problem-solving of human brain is the same as the structure of the symbolic logic system of computer because computer learns via trial and error like a human. The difficulty in that time was the creation of productive Artificial Intelligence for non-programmed decision-making. After that, though Public Administration was in the third paradigm and the fourth paradigm from 1950 to 1970, AI as a part of computer science is in the writings of PA scholars in the fourth paradigm such as Simon, Cyert and March. (Frederickson, 1980; Henry, 1975, 2010) In the fourth paradigm, before Cyert and March s writing in 1963, Simon with Newell and Shaw (1958) firstly argued that decision-making process is in terms of a program by using digital computers. They stated that We wish to emphasize that we are not using the computer as a crude analogy to human behavior we are not comparing computer structures with brains, nor electrical relays with synapses. Our position is that the appropriate way to describe a piece of problem-solving behavior is in terms of a program: a specification of what the organism will do under varying environmental circumstances in terms of certain elementary information processes it is capable of performing. This assertion has nothing to do directly with computers. Such programs could be written (now that we have discovered how to do it) if computers had never existed. A program is no more, and no less, an.analogy to the behavior of an organism than is a differential equation to the behavior of the electrical circuit it describes. (Newell, Shaw and Simon, 1958: 153) Later, Simon, Smithburg, and Thompson published their writing in 1961, Public Administration, did not directly present anything about AI in their book. However, they suggested intelligence units for gathering information for organizational use from both internal and external sources. In addition, they also classified these intelligence unit into two types: external intelligence unit for gathering external information for internal use, and internal intelligence unit for collecting information for controlling organization. (Simon, Smithburg and Thompson, 1961) After that, in 1963, Cyert and March, in their A Bahavioral Theory of the Firm, first pointed that the process of decision-making is in the form of flowcharts, algorithms, computer programs (Cyert and March, 1963), which is the emerging points of the digital computer in Public Administration. Two years later after Cyert and March, Simon indicated in his writing in 1965, The Shape of Automation for Men and Management, that heuristic programming, which was called Artificial Intelligence (AI) in the third edition of the 1965 textbook-the new science of management decision-in 1977, approach an automated solution to solve non-programmed [11]

5 problems by using experience-based rules. In addition, Simon also distinguished Operation Research (OR) for programmed decisions from heuristic programming for non-programmed decisions. (Simon, 1965, 1977) After that, 3 years later, PA has been under the influence of the Minnowbrook s doctrine in the fifth paradigm (PA as PA since ) and the sixth paradigm (Governance since 1990 until now). In this period, there are lots of works about AI as vital tools for policy analysis and knowledge management (KM). The formal beginning of AI formally started in 1972 when Simon and Newell drew up a manifesto for AI in 1972, Human Problem Solving, by stating that AI is the part of computer science that focuses on the replacement of human brain with computer. This book is full of problem-solving researches within an explicit framework of computer science theory for two decades of Simon (Simon and Newell, 1972; Simon, 1979). After that, a decade later, Simon and Newell, also repeated the difference between OR and heuristic programming in 1982, Model of Bounded Rationality: Volume 1 Economic analysis and Public Policy, by pointing that OR is for well-structured problems and heuristic problem solving is for ill-structure problems (Simon, 1982). Later, Simon, in his Reason in Human Affairs in 1983, directly indicated AI as one of technical tools for decision-making and public policy by pointing that AI is for qualitative consideration whereas operation research is for mathematized decision (Simon, 1983). In addition, Simon, Smithburg, and Thompson in their new edition of Public Administration in 1991 indicate the importance of IT for PA and public policy (Simon, Smithburg, and Thompson, 1991). After that, in 1999, Barth and Arnold in their article, Artificial Intelligence and Administrative Discretion: Implications for Public Administration, describe both the advantages and disadvantages of AI for PA (Barth and Arnold, 1999). However, in the Minnowbrook period with three conferences in 1968, 1988, and 2008 for investigating the state of the art and trend of PA, AI is not one of the important issues for PA in the study of the provision of public goods and services in all three Minnowbrook conferences. In Minnowbrook I, AI was indirectly discussed as a tool for data respiratory for KM, Public Administration Theoretical Inventory (PATI) on punch cards, and as a system approach of Planning-Programming-Budgeting System (PPBS) (Marini, 1968). Next, the PA scholars in Minnowbrook II in 1988 poorly addressed the importance of IT such as electronic mail, ETHNET, SOCNET, POLINET, COMPUSOURCE, word processing, and AI as a vital drive for public reform (Guy, 1989). After that, in Minnowbrook III, the application of AIbased tools for public policy analysis such as process models and social network analysis are discussed less than 1 page as analytical tools for dealing with complex problems in collaborative Governance and cross-boundary information sharing (O Leary, Van Slyke, and Kim, 2010). One reason might be the legacy of unfamiliarity and uncomfortability with social psychology and IT since paradigm III (Henry, 1975, 2010). Conclusion This article is the study of AI application for PA from many key books of PA scholars since Though the findings found that AI is still not recognized by PA scholars in three Minnowbrook conferences as an important tool for public policy analysis and e-government, it has a big issue for further study. References Barth, T. & Arnold, E Artificial Intelligence and Administrative Discretion: Implications for Public Administration. American Review of Public Administration 29 (4): [12]

6 Bowornwathana, B Public Administration: Theories and Methodologies. Bangkok: Chulalongkorn University Press. Cyert, R. & March, J A Behavioral Theory of the Firm. New Jersey: Prentice-Hall. Dervojeda et al Big Data Artificial Intelligence (Business Innovation Observatory Contract No 190/PP/ENT/CIP/12/C/N03C01). Retrieved June 6, 2015 from ec.europa.eu/enterprise/policies/innovation/policy/business-innovation-observatory/ files/case-studies/09-bid-artificial-intelligence_en.pdf. Frederickson, G New Public Administration. Alabama: University of Alabama Press. Goodnow, F Politics and Administration: A Study in Government. London: Macmillan. Guy, M Minnowbrook Ii: Conclusions. Public Administration Review 49(2): Henry, N Paradigms of Public Administration. Public Administration Review 35 (4): Henry, N Public Administration and Public Affairs. New York: Longman. Laudon, K. & Laudon, J Management Information Systems: Managing the Digital Firm. Essex: Pearson Education, Marini, F. (ed.) Toward a New Public Administration: The Minnowbrook Perspective. New York: Chandler. Newell, A., Shaw, C. & Simon, H Elements of a Theory of Human Problem Solving. Psychological Review 65 (3): PAR PAR 75 Most Influential Articles. Retrieved June 6, 2015 from ministrationreview.org/full-list/. Pardo, T., Gil-Garcia, R. & Luna-Reyes, L Collaborative Governance and Cross- Boundary Information Sharing. in R. O leary, D. Van Slyke and S. Kim (eds.). The Future of Public Administration around the World: The Minnowbrook Perspective. Washington D.C.: Georgetown University Press. Robbins, S., Decenzo, D. & Coulter, M Fundamentals of Management: Essential Concepts and Applications. Essex: Pearson Education Limited. Shafritz, J. & Hyde, A Classics of Public Administration. Texas: Harcourt Brace College Publishers. Schermerhorn, Jr., J Introduction to Management. New Jersey: John Wileys & Sons. Simon, H Administrative Behavior. New York: Free Press The Shape of Automation for Men and Management. New York: Harper & Row The New Science of Management Decision. New Jersey: Prentice-Hall Models of Thought. London: Yale University Reason in Human Affairs. Oxford: Basil Blackwell Polisher The Sciences of the Artificial. Cambridge, Ma: The Mit Press. Simon, H. & Newell, A Human Problem Solving. New Jersey: Prentice-Hall Heuristic Problem Solving: The Next Advance in Operation Research. in H. Simon (ed.). Model of Bounded Rationality: Volume 1 Economic Analysis and Public Policy. Massachusetts: MIT Press. Simon, H., Smithburg, D. & Thompson, V Public Administration. New York: Alfred A. Knopf Public Administration. New Jersey: Transaction Publishers. Urwick, L The Function of Administration with Special Reference to the Work of Henri Fayol (1934). in Gulick, L. and Urwick, L. (eds.). Papers on the Science of Administration. New York: Institute of Public Administration, Weber, M From Max Weber: Essays in Sociology. New York: Free Press. [13]

7 Theory of Social and Economic Organization. New York: Free Press. Wilson, W The Study of Administration. Political Science Quarterly 2: Yick, L Organizing around Intelligence: The New Paradigm. Singapore: World Scientific Publishing. [14]