How Does Problem Analysis Determine the Quality of Policy Decisions?
“Defining the problem correctly is half the solution” is more than a well-known aphorism. It is recognized as an important methodological principle in modern policymaking. The European Commission’s Better Regulation Toolbox presents problem analysis as one of the first and most critical stages of impact assessment. It emphasizes that only an accurate diagnosis of a problem and its causes makes it possible to select appropriate policy measures. The OECD likewise notes that how a problem is understood and defined shapes the direction of the entire subsequent policy cycle. The quality of a government program should therefore be assessed not only by the number of measures it contains or the ambition of its targets, but also by the diagnosis of the problem from which those measures originate.
This principle is particularly important in government programs. An error in an initial diagnosis also affects the subsequent selection of objectives, measures, subsidy mechanisms, and outcome indicators. In other words, if the problem is not properly defined, the government may later allocate resources to solving the wrong problem. In such a case, even if the program consists of measures that appear logical on paper, its ability to justify the need for government intervention and substantiate the expected outcomes is weakened.
Approved approximately one month ago by a relevant Order of the head of state, the “State Program for the Development of Tourism in the Republic of Azerbaijan for 2026–2030” provides an interesting case for examining this issue. The Program has broad objectives: strengthening institutional capacity, stimulating tourism demand, developing tourism supply, improving service quality, and strengthening the information base for decision-making. It also seeks to improve the investment climate, attract private investment, and establish various incentive mechanisms.
However, this raises a central issue that determines the quality of policy decisions: The Program does not establish a sufficiently clear link between its targets and certain financial interventions, on the one hand, and the diagnosis of the problems on which they are based, on the other. This article presents two particularly clear examples.
Example I: Subsidies Without Adequate Justification
The Program envisages subsidy mechanisms in various areas to support tourism development, yet almost none of them is supported by a rigorous justification. For example, the Program proposes subsidizing the construction of new hotel rooms.
The diagnostic section of the Program (paragraph 3.2.4) states that Azerbaijan had 0,25 hotel rooms per 100 people in 2021, which was 2,6 times below the global average. It then states that improving the investment climate will increase the adequacy of supply and optimize prices. Measure 8.8.2.4 of the Program’s action plan subsequently provides for the preparation and adoption of a normative legal act introducing a subsidy for each newly built and commissioned hotel room, followed by the launch of the subsidy mechanism. The outcome indicator for measure 8.8.2.4 is defined simply as “an increase in the one-time capacity of accommodation establishments.”
This raises a significant methodological problem. The fact that there are 0.25 hotel rooms per 100 residents does not demonstrate that existing hotel capacity is insufficient to meet demand. The figure cited in the Program is merely an international benchmark showing that supply capacity is low relative to the population. Yet demand for hotel capacity in the tourism sector is not determined by the size of the country’s population. It is therefore inappropriate to use this figure as evidence of a hotel-room shortage in the country’s tourism market.
A hotel-room shortage in the tourism market is a market problem specific to a particular destination, season, and demand structure. The central question is: What proportion of actual tourist demand can the existing hotel-room supply accommodate during peak and average periods? Answering this question would have required, at a minimum, the following indicators:
- the number of existing hotel rooms;
- the annual growth rate in room supply;
- the hotel occupancy rate;
- monthly and quarterly occupancy rates;
- the occupancy rate during the peak season;
- occupancy rates by region;
- the specific destinations experiencing room shortages;
- whether tourists encounter difficulties in finding hotel accommodation;
- rejected reservations and incidents of “room shortages”;
- the response of hotel prices to growth in demand;
- the average daily room rate (ADR) and revenue per available room (RevPAR);
- the proportion of the existing room inventory that is actually in use; and
- projected tourist arrivals relative to existing capacity.
The Program, however, provides no analysis of this kind of supply-demand balance. To justify the subsidy, an assessment should have been conducted using actual and peak-season occupancy rates, destination-level supply-demand gaps, unmet demand, price effects, and other relevant indicators. Most importantly, the Program should have answered the question: Why is the market not building enough hotel rooms on its own?
To answer this question and justify government intervention, the Program should have identified a market failure or a specific barrier to investment. Potential barriers include high capital costs, long investment payback periods, land and infrastructure constraints, interest rates, low seasonal profitability in the regions, the high risks associated with hotel investment, utility and infrastructure costs, and legal and administrative procedures. Although the Program refers to some of these issues in general terms, it does not quantitatively assess which barriers reduce the attractiveness of hotel investment.
The Program also presents another noteworthy fact: Seasonality in the country’s tourism sector is very high, and 51,2% of accommodation revenue is generated during just five months of the year. This complicates the case for subsidizing an expansion in hotel-room supply. If the existing hotel stock remains underutilized for a substantial part of the year, the efficiency with which existing rooms are used must be assessed before new rooms are built.
Finally, another noteworthy point is that the outcome indicator for the subsidy measure is defined as an “increase in hotels’ one-time capacity.” Yet an increase in the number of rooms does not necessarily imply their economically efficient use, the generation of additional tourist demand, or an increase in local value added.
The Program therefore fails to demonstrate the necessity of government subsidies for the construction of new hotels. Instead, we encounter the problem that the European Commission strongly warns against: “backward engineering”—that is, working backwards to justify a solution selected in advance.
The same reasoning applies to the Program’s other subsidy proposals. The Program provides mechanisms to compensate part of the costs of events of significance to tourism, subsidize investors in accommodation infrastructure, subsidize tour operators for bringing foreign tourists into the country, and provide subsidies to micro and small enterprises.
Each of these instruments requires its own economic justification. For an event subsidy, it must be established whether the event would take place without government support and whether it would generate net additional tourist demand. For a tour-operator subsidy, the analysis must determine how many of the subsidized tourists would not have travelled to Azerbaijan in the absence of government support. For subsidies to micro and small enterprises, it must identify the specific financing constraint that the market does not address. Otherwise, an “increase in activity” may be conflated with an “additional outcome generated by government intervention.”
Example II: The target increase in tourist numbers is not broken down by contributing factors.
The Program aims to increase the number of foreign nationals and stateless persons visiting Azerbaijan from 2,57 million in 2025 to 3,8 million in 2030. It also envisages increasing the number of domestic tourism trips from 5,9 million to 7,4 million.
These figures are important outcome targets. The key issue, however, is that the Program does not specify in advance which factors are expected to drive this growth. Tourist flows may increase because of natural market growth, expanded air connectivity, marketing, new tourism products, entry conditions, the operating regime of land borders, international events, visits by Azerbaijanis living abroad and by people visiting relatives, as well as government subsidies. It would be methodologically unsound to attribute subsequently all growth arising from these sources to the State Program.
For example, not every foreign national entering the country is automatically a tourist with the same economic profile. Visitors may be classified according to their economic characteristics as leisure tourists, business travelers, people visiting relatives or friends, event participants, and transit passengers. These groups may differ substantially in their length of stay, daily expenditure, use of hotel services, travel to the regions, and demand for tourism products. The aggregate indicator of the “number of foreign arrivals” therefore does not fully capture actual market demand in the tourism sector.
It is also important to assess and account for the role that events organized at the national level and by government bodies—such as COP29, Formula 1, conferences, and festivals—play in tourist flows. The Program treats international events and events of significance to tourism as positive factors and envisages organizing further events, including by providing subsidies to event organizers. For example, measure 6.1.8.2.2.3 provides for a subsidy mechanism to reimburse part of the costs of concerts, entertainment programs, and other events of significance to tourism. One of the expected outcomes is an increase in the number of large-scale international events held in Azerbaijan. However, there is a serious problem: The Program draws no analytical distinction between holding an event and the net additional tourist inflow generated by that event.
Another issue is the continued closure of the land borders. The diagnostic section of the Program refers in general terms to the effects of the restrictions imposed during the COVID-19 pandemic and the continued closure of the land borders on the tourism sector. However, it does not quantify the impact of these factors on tourist flows. Specifically, the Program does not answer the following questions:
- How many additional tourists might have arrived if the land borders had been open?
- How many people entered the country through land borders in 2019, and what share of total arrivals did they represent?
- How many potential trips were prevented by the closure of this entry channel between 2020 and 2025?
- What proportion of tourists arriving from Russia, Georgia, Iran, and Türkiye use land transport?
- How much could the reopening of the land borders contribute to growth in tourist arrivals in 2026, 2027, and subsequent years?
- What proportion of the projected increase from 2,57 million to 3,80 million visitors by 2030 can be explained by a change in the border regime?
Thus, a potential change in the land-border regime—one of the key determinants of the growth trajectory for 2026–2030—is not addressed, at least not in the Program itself. This is a serious omission. The border regime is an exogenous factor controlled by the government. If a visitor target is set for 2030, the impact of such an important factor—one that the government can alter through its own decision—should be modeled as a separate scenario.
This gap is not merely statistical. It affects the quality of the policy decisions made at the outset of the Program and also impedes the subsequent evaluation of its results. Unless the sources of growth in tourist arrivals are decomposed in advance, any increase recorded by 2030 cannot be reliably attributed to the measures under the State Program, the reopening of the land borders, government-organized events, general market growth, or other external factors.
Conclusion
The quality of a government program is measured less by the number of measures it contains than by the diagnosis from which those measures are derived. Defining the problem correctly is the starting point of policy; correctly identifying its cause determines which form of government intervention is genuinely necessary. In tourism policy, the central question should not simply be: How can we attract more tourists and expand hotel-room capacity? The essential question is: What specific problem is constraining tourism development, what is causing it, and—by addressing that cause—what additional outcome will government intervention produce that the market could not generate on its own?
