PTStrategy, technology and performance in city halls
A city hall can innovate, defend what it already does well or just react. It can spend time and paper on every reply. And it can have more or less technology. With 131 civil servants from city halls in inland Paraná, Brazil, the thesis asks which of these goes hand in hand with better public service. Explained in 12 screens.
Answering the city council once took up to 30 days of back and forth
In one of the study’s interviews, a civil servant describes how it used to work: each request from the city council arrived on paper, was copied for each department and went back and forth with the mail driver. Finding the information, chasing whoever had not replied, checking: theory calls these transaction costs, costs that do not show up in the budget. Compare:
Our illustration based on the account in Interview 02 (Appendix C). The types of cost (search, negotiation, monitoring) come from the thesis introduction. One interviewee sums it up: these are costs “a bit harder to measure” (Interview 01).
Miles and Snow describe four ways of dealing with change
In 1978, Miles and Snow proposed four strategic postures: the prospector seeks the new, the defender protects what it already does well, the analyzer balances the two and the reactor only moves when pushed. Tap a posture to see how the thesis measured it and how interviewees described themselves:
Descriptions from section 2.1.4 and Figure 1 (adapted from Miles and Snow, 1978). In the analysis, the thesis merged prospector and analyzer into a single posture. Means from 1 to 6 from Table 3. Quotes from Appendix C, by interview number.
Monitoring was the part of transaction costs with the most agreement
Coase and Williamson’s theory explains why every exchange costs something beyond the price. The thesis grouped the questions into three parts: bounded rationality, asset specificity and monitoring. Tap a column to see what each one means and the statements civil servants rated:
Means from 1 (strongly disagree) to 6 (strongly agree), Table 1. The parts came from a factor analysis. Our note: the statements ask for agreement with ideas from the theory (“transaction costs are a reality”), not the amount spent.
131 civil servants from city halls in inland Paraná answered
First, the author talked to civil servants from departments in Foz do Iguaçu to understand how they speak about these topics. With that, he built a questionnaire, tested it and shortened it. The link went to staff in the Administration and Information Technology departments and the Mayor’s Office, who needed some basic knowledge of the subject.
Sections 3.1 and 3.2. Convenience sample; answers from 1 to 6, from strongly disagree to strongly agree. The pre-test also cut 6 minutes from the answering time. Analysis with structural equation modelling (PLS-SEM, SmartPLS 4).
Percentages from section 4.2; people per group = percentage × 131, rounded (our calculation).
Answer as a civil servant
Six statements from the questionnaire, as summarised in the thesis tables, from 1 to 6. Think of the city hall or public body you know. The mean of the 131 appears after your answer.
The bar starts at 1. The postures measure how important the civil servant thinks each attitude is in the city hall’s planning (“the city hall’s capability to… is important”); performance is the perception of the last three years (Appendix D).
Items PL71, PL81, CT1, INFRA3, CAPA3 and DO1; means from Tables 1 to 4.
For civil servants, innovating and maintaining almost merge
In theory, the prospector posture (seeking the new) and the defender posture (keeping what works) are opposites. Say which posture each questionnaire statement belongs to:
Correlation between the two postures: 0.912 (Table 5); HTMT of 0.955, above the 0.9 limit the thesis itself uses. Circle overlap ≈ r² (about 83%, our calculation, illustration). Means from Table 3.
Three arrows reach performance, and two come from technology
The thesis tested everything at once (structural equations). In the final model, the reactor and prospector postures are linked to the parts of transaction costs, the defender is linked to technology, and performance receives arrows from infrastructure, IT capabilities and the reactor. Tap an arrow:
Final model from Figure 5 and Table 7 (SmartPLS 4): standardised coefficients (β), thickness proportional to β. The three parts of transaction costs (bounded rationality, asset specificity and monitoring) appear together in one block; tap the arrow to see each. None of them has an arrow to performance in the final model. R² = how much of the differences between answers the model explains.
The defender posture reaches performance through technology
Those who value keeping services running and protecting citizens also rate IT infrastructure and capabilities higher, and it is through them that the defender posture reaches performance. The prospector does not. Together with the reactor, the model explains 61% of the differences in perceived performance. Switch the view:
Height = size of the standardised effect. Through infrastructure: 0.469 × 0.382 = 0.179; through capabilities: 0.457 × 0.328 = 0.150; together 0.329, all significant (Table 7). Grey and flat = no clear effect (Table 8). The cup goes up to 100%.
Only the reactor posture kept a direct link to performance
In theory, the reactor is the posture with no plan, the least competitive. Before you look, a guess: which posture do you think kept a direct link to performance?
Go together (simple correlation)
Direct link in the model (β)
Correlations with performance from Table 5; direct links from Table 8 (prospector 0.001, p 0.988; defender −0.040, p 0.640) and Table 7 (reactor 0.267, p < 0.001). Bars up to 0.6.
Ten of the 23 links tested were confirmed
Each posture was tested against the three parts of transaction costs, infrastructure, IT capabilities and performance; then each of these pieces against performance. The parts of transaction costs respond to the postures but hardly reach performance: only a hint, in asset specificity. Filter and tap a block or a number in the grid:
Table 8 (23 hypotheses) and, for the significant ones, Table 7 and Figure 5. Height = β, without the sign. Solid colour = confirmed at 5%; striped = hint (p between 0.05 and 0.10, which the thesis counts as confirmed); grey = no clear effect.
Invest in technology and start measuring
For the author, the study can serve as a basis for administrative modernisation policies in inland cities. The interviews and the limits the author points out show the missing step: indicators to measure one’s own performance.
For city halls
- IT infrastructure and capabilities were the largest links to perceived performance (0.382 and 0.328).
- Training staff and having robust systems are, for the author, critical to these policies; civil servants agree IT requires investing in both (4.42 and 4.41).
For public managers
- “Improve a performance indicator? We don’t have that kind of tracking.” (Interview 02)
- Our reading: without indicators, performance can only be measured as people’s impression, as in this study.
For strategy researchers
- In the public sector, postures blend: prospector and defender scores barely separate, as Boyne and Walker, cited in the thesis, had argued.
- The author suggests studies in other regions and over time.
What this study does not say
- Small convenience sample. 131 civil servants from inland Paraná; other regions may differ. · according to the author
- Few indicators in city halls. The lack of clear indicators made it harder to assess the impact of technology. · according to the author
- Perception, not measurement. Performance is what the civil servant perceives over the last three years, and the postures are how important each attitude seems to them, not the city hall’s type. · our note
- A single moment. Everything was asked in the same questionnaire: it shows what goes together, not what causes what. · our note
- Two postures that do not separate. Prospector and defender were above the similarity limit; the author kept them apart to preserve the typology. · Table 5
Credits
STOECKL, Karl. Efeitos da infraestrutura e capabilidades de TI, a tipologia de Miles & Snow e o custo de transação no desempenho operacional, no contexto de Gestão Pública. Doctoral thesis (Business Administration), Mackenzie Presbyterian University, São Paulo, 2024. Advisor: Prof. Dr. Silvio Popadiuk.
Thesis approved on 16 December 2024. The author thanks his first assigned advisor, Prof. Dr. Roberto Giro Moori, and União Dinâmica de Faculdades Cataratas (UDC) for institutional support; the research was part of an inter-institutional doctorate (DINTER). Interviews analysed in ATLAS.ti and the questionnaire in SmartPLS 4. This version simplifies the work for the general public.