The Nigerian Statistical System is very disjointed. Major producers of data include National Bureau of Statistics and the Central Bank of Nigeria. The result is disharmony in data collection and publication. Data under the same heading are often inconsistent and disjointed. Coordination as in Canada is the best scenario because this reduces discrepancies between data produced by different bodies. 

Ette - Thank you.  That is interesting.  That seems to me to be the same chaotic result we see in the United States. 

In addition, in some cases, perhaps one way that it may help to have a national statistical system would be to reduce redundancy of survey data collections.  (Redundacy might be helpful for data reliability checking, but not if one is not allowed to examine all surveys.) Also the ability to provide administrative data for use as auxiliary or regressor data elsewhere, may further improve overall efficiency.  It is difficult to obtain useful administrative data in the United States, where we do not have such cooperation. 

US is one of the few official statistical systems that are fragmented. In most countries there is a national institute that does most of the official statistics work. A special care is Canada, where there are two official statistical institutes, Statistics Canada and the Quebec Statistical Institute, in which case there is some duplication. In some cases it is a good thing, because if one institute has a poor-performing division, the similar division in the other institute may be really good. Another interesting thing in Canada, that could be true for other federal states, is that provinces take an active role in supplementing gaps in data through their statistical entities in order to produce reliable statistics.

A major advantage of having a decentralized official statistics system is that it is accountable to the organization it belongs to and benefits from some professional scrutiny may not happen in centralized institutes. On the other hand, the decentralized units may be(come) prisoners of vested interests that are helped through dissemination and use of statistics.

Just a few thoughs, maybe people who know more will weight in...

Please let me know if the following references/sites are helpful to you on your quest:

1.  Vital statistics systems - OECD.org

http://www.oecd.org/site/worldforum06/38756222.pdfThe United Nations Statistical Commission and the United Nations Statistics Division have ... Office of the United Nations Secretariat, which included assuming the statistical work of the ... their own national standards for official statistics. ..... Standardization of Vital Statistics, developing countries still do not have complete ...

2.  United Nations Fundamental Principles of Official Statistics - UNSD

https://unstats.un.org/unsd/dnss/gp/Implementation_Guidelines_FINAL_without_edit.pdfUN Fundamental Principles of Official Statistics – Implementation guidelines, 2015. 2. Table of ... The United Nations Statistics Division remains committed to working with all coun- ... a high level of independence of national statistical systems. .... (http://www.turkstat.gov.tr/UstMenu/yonetmelikler/rip.pdf) and

3.  Models of Statistical Systems - Paris21

https://www.paris21.org/sites/default/files/2101.pdf

United Nations Statistics Division, 'Fundamental Principles of Official Statistics,' ... principles do not define a national statistical system, although a working.

Dennis

Dennis Mazur

The more groups determining the statistics the better. Adrian is right; one group controlling all of the statistics can lead to manipulation. Have a large number of independent groups getting the data is incredible redundancy, but safer for a free society. Heck, even elementary statistics tells us that one observation is not enough to determine what's going on. One group is one observation.

Very good point, Adrian and Peter, regarding one statistical program checking another.  (I was not aware of the details on Canada which you provided, Adrian. Thank you.)  But in the US, agencies generally are not allowed to share data, so if data from one agency might inform another regarding issues such as regressor/auxiliary data, or even enough information to be better able to stratify when sampling, it may be difficult to impossible to allow  cooperation.  If say a manufacturers' energy  consumption survey would benefit/be more efficient if the survey methodology were planned and estimators based on supplementary information, such as number of employees or expenditures, which may establish a size measure for each member of the population, that kind of thing may not be possible.  US agencies use memoranda of understandings, MOUs, to try to work together when allowed, but even then it can be tedious/inefficient.  

James, look at the equation of exchange. Velocity is usually the "leftover" variable after the price level, money supply and real GDP have been determined. But what if we found velocity thru surveys and let GDP be the leftover? Likely, we'd have slightly different results; maybe better, who knows. Getting data one way is efficient but scary.

 I'm actually thinking of model-assisted design-based sampling and estimation, and strictly model-based sampling and 'prediction' (not forecasting), where I've done a lot of work in the latter, used by the US Energy Information Administration.  Looking quickly at a list of many of the books I own, I noticed the following would qualify as having techniques requiring auxiliary/regressor data, if any are familiar to you: 

Särndal, CE, Swensson, B. and Wretman, J. (1992), Model Assisted Survey Sampling, Springer-Verlang.    

Brewer, KRW (2002), Combined survey sampling inference: Weighing Basu's elephants, Arnold: London and Oxford University Press.  

An Introduction to Model-Based Survey Sampling with Applications, 2012,

Ray Chambers and Robert Clark,  Oxford Statistical Science Series

Finite Population Sampling and Inference: A Prediction Approach,  2000,

Richard Valliant, Alan H. Dorfman, Richard M. Royall,

Wiley Series in Probability and Statistics.

Survey Sampling: Theory and Methods, First Edition, 1992,

Chaudhuri, A., Stenger, H.,

Marcel Dekker, lnc., New York, Basel, Hong Kong.

Thompson, S.K.(2012), Sampling, 3rd ed, John Wiley & Sons.

Cochran, W.G(1977), Sampling Techniques, 3rd ed., John Wiley & Sons. 

Blair, E. and Blair, J(2015), Applied Survey Sampling, Sage Publications.

Lohr, S.L(2010), Sampling: Design and Analysis, 2nd ed., Brooks/Cole. 

Actually, my older books on generally design-based methods still include ratio and regression estimation most if not all of the time. 

This book also notes ratio estimation:

Snijkers, G., Haraldsen, G., Jones, J., and Willimack, D.K.(2013), Designing and Conducting Business Surveys,  John Wiley & Sons, Inc. 

Here is an excellent paper which succinctly explains the difference between strictly design-based methods and the other two, where auxiliary/regressor data are needed:

===================

Ken Brewer's Waksberg Award article: 

 Brewer, K.R.W. (2014), “Three controversies in the history of survey sampling,” Survey Methodology,

(December 2013/January 2014), Vol 39, No 2, pp. 249-262. Statistics Canada, Catalogue No. 12-001-X.

http://www.statcan.gc.ca/pub/12-001-x/2013002/article/11883-eng.htm

He believed in using probability sampling and models together, but he explains the different approaches - the pros and cons. 

===================

Oh yeah.  I have this book, for which I wrote a book review for the Journal of Official Statistics, which I recall says a good deal about the importance of being able to obtain administrative data for auxiliary data, if I remember correctly:

Estimation in Surveys with Nonresponse, Carl-Erik Sarndal and Sixten Lundstrom, John Wiley & Sons, 2005.

That stresses "calibration weights," which means reweighting survey weights in estimation, based on auxiliary data to obtain generally more accurate results with less new sample data collection. 

Strictly model-based ratio estimation  ('prediction') for many small populations with small samples under circumstances found with data collected at the US Energy Information Administration (EIA) is greatly aided by the regressor data for ratio estimation.  Fortunately, those regressor data are also collected by the EIA.  There are simple cases of say an annual census which could provide regressor data for a monthly sample. 

.....

I think that the Lundstrom and Särndal(2005) book noted above is probably potentially more useful when there is a national statistical agency, or a national system which shares data well, rather than a nation such as the US, where it's techniques appear to not be so likely feasible to apply. 

Note that this is not just about nonresponse.  Calibration works very well when nonresponse is not the issue.