5.0 Answering the key questions

As a result of the research done, the Agriculture Division is able to respond to the three questions asked at the outset of this review.

1) Is a CEAG still the best way to meet the data requirements for policy and program purposes? If so, what should its frequency be? More specifically, is a CEAG required in 2016?

The review confirmed that a complete enumeration of the agriculture industry (CEAG) is needed to meet the policy and program requirements of federal and provincial governments, industry and other key stakeholders. The activities conducted by the federal and provincial governments that depend on this completeness are numerous, including health policy, land planning, crisis management, international trade, environmental accounting and reporting, and global commitments to the international community. Several of these inter-disciplinary activities are massive undertakings spanning more than one federal department as well as provincial governments. However, it is not solely the agriculture industry that relies on data from the CEAG. In the absence of another source of data that fully enumerates the industry, a CEAG is still the best way to meet these requirements.

The quinquennial CEAG data required to re-align survey estimates and survey frames are also critical to Statistics Canada, other federal departments (particularly AAFC) and provincial governments. At this time, no other source of information exists from which to extract this information. Intercensal frame deterioration is a current challenge despite the fact that a CEAG is conducted every five years. The magnitude of the intercensal revisions can sometimes be significant as a result. For policy and program evaluation and performance reporting to Treasury Board, re-alignment of the estimates decennially is insufficient, as it can lead to programs that are out of alignment for a long period of time. Accurate estimates are especially important given the volatility in the industry and the level of support that governments disburse to the agri-food industry.

Through consultations with key data users and through the assessment of the requirements for agriculture statistics thus far, it has become evident that a CEAG conducted less frequently than every five years will result in data gaps that could not be filled by any other means in the short and medium term. Preliminary discussions with the major stakeholders revealed that they are unprepared for these data gaps. In the absence of the 2016 CEAG, the work performed by several federal and provincial departments would be impacted due to the cross-sectional nature of many policies and programs. The volatility in the industry discussed earlier in the report further raises the need for a quinquennial enumeration of the industry. The most affected external departments would be AAFC, Health Canada, Environment Canada and the provincial governments. The absence of the 2016 CEAG data would also impact the entire agriculture statistics program, due to its integrated nature. These reasons support conducting a CEAG in 2016.

2) Given the data requirements for policy purposes, is the CEAG in its current form the most efficient way to gather the information, and are there efficiencies to be gained in the CEAG?

There are efficiencies that could be gained in the CEAG over time that could provide for the requirements of complete enumeration and survey re-alignment. Several features have been identified that could reduce some of the burden and cost associated with the current CEAG, while continuing to satisfy the requirements of complete enumeration and survey re-alignment that are so critical on a quinquennial basis.

The chosen option must take into consideration the requirements for policy and program development, monitoring and evaluation by collecting the required information from all agriculture producers, but limiting the CEAG content to these specific requirements.²⁴ Regardless of the chosen option, eliminating questions that can be replaced with taxation data or other administrative data could reduce the CEAG content considerably over time.

3) How can the agriculture statistics program as a whole be streamlined to reduce response burden and costs, while continuing to meet priority data requirements?

The current Canadian agriculture statistics program comprises a quinquennial CEAG, the commodity-specific surveys, administrative and taxation data, remote sensing, the agriculture economic statistics series as well as the research and analysis work. These components can be considered bricks that constitute the foundation of the program.

Taking a progressive approach to integrating new features into the current Canadian system reduces the risk of incurring significant investments in an entirely revised program and also reduces the risks of error and loss of coherence associated with more substantial changes. By maintaining the foundation of the current program, the basic structure could remain intact while being adapted over a period of several years. In this way, the strategies of replacing survey data with administrative data, taxation data and remote sensing technology will result in efficiencies, while minimizing the risks to the relevance, coherence and accuracy of the program.

Respondent burden could be reduced over time as new administrative data sources are identified, evaluated and incorporated into the agriculture statistics program. Remote sensing also has the potential to play a more important role in supporting the agriculture statistics program. Further work will be required to quantify the investments, savings and timelines associated with the adoption of administrative data and remote sensing technology.

Survey response burden could be alleviated by reducing either the target population or the survey population. Reducing the target population would affect the coherence and comparability of the data, whereas maintaining the same target population and reducing the number of farms eligible to be surveyed could allow the currently published estimates to be maintained. The quinquennial CEAG provides regular data for the modelling of the non-surveyed population. This strategy requires no investment to adjust historical data to a new target population definition.

Further cost savings could also be introduced into the program by rationalizing and reducing the number of survey occasions per year for some crop, horticulture and livestock surveys where user data requirements can continue to be met. In addition, further response co-ordination in the CEAG years could reduce response burden over time as other sources of data become incorporated into the program.

It is possible to continue to develop a revised agriculture statistics program that respects the international agriculture statistics priorities and guidelines, namely:

• the recent commitments made by the Agriculture Ministers of the G20 regarding improvements to agriculture statistics and the Agriculture Market Information System (AMIS)
• alignment with the global initiative to improve agriculture statistics, as described by the FAO²⁵
• alignment with the World Bank and FAO's Global Strategy to Improve Agricultural and Rural Statistics.²⁶

Several of the features presented in the various options could result in further opportunities for reducing response burden and finding cost efficiencies. Increasing the utilization of remote sensing, increasing the incorporation of taxation data and administrative data will result in reduced response burden, realized cost efficiencies and a good quality program in the medium to longer term.