2nd Berkeley Meeting, Jan 2005
SUMMARY OF THE MEETING
Lifecycle Deficit
- Public consumption
How should public investment be treated. Two options: 1) count as consumption; 2) track investment with consumption equal to depreciation and the return to capital.
Suggestion: Use one country as a test case. Any volunteers?
Suggestion: For first year of the project: simply count government spending as consumption. In subsequent years: we can introduce more detailed modeling of the public sector exploring the issue of public investment. We will need to track government investments in the different sectors: health, education, environment, etc. -- each with their own depreciation rate. Also raises issues of how to value the environment as an asset. This is hard but important work -- best left out of our first year plans. -[tm]
- Private consumption
Methods for allocating household consumption among individuals are relatively unreliable. Consensus of group is to rely on direct allocation. Education and health would be allocated using regression methods. All other consumption is allocated using an equivalence scale common to all countries. A working group will draft a policy. The group is Bommier, SH Lee, Maliki, Mason, Miller, Wolff.
- Labor income
In some labor markets, current wages may diverge from current productivity. Seniority wage system in Japan is a good example. Treatment depends on whether one believes that this is a transfer (from young workers to old workers) or whether young workers are essentially paying for their own higher wages. Also, the treatment depends on whether or not the deferred wages are funded.
Suggestion: SH Lee, Riki, and Hiro consider this issue and make a recommendation.
Asset reallocations
- Simplification by merging capital, credit, and land
- Financial intermediaries
Briefly discussed. Two approaches were discussed. The first would provide more detailed tracking of financial transactions. The second approach would classify saving between capital and credit transactions based on the end use of the funds.
Suggestion: Mason will work on this.
- Treatment of SOEs
General and relatively wide-ranging discussion of this issue. One issue is whether to classify investment in SOEs as private or public investment. Another issue is how to deal with the non-competitive environment in which public firms may operate. If prices to consumers do not reflect marginal cost, then it may be important to estimate the extent to which consumers are being taxed or receiving transfers. Similarly, wages paid to employees may include subsidies or not.
Suggestion: An-Chi will work on this.
- Public investment
See public consumption above.
Suggestion:
Land and Credit Reallocations
Public Transfers
- Which public programs?
At a minimum we want to distinguish health and education from other public programs. The unresolved issue is how much more detail than this is possible.
Suggestion: One approach is to set a goal for the first year: education and health analyzed in detail; all other allocated using simple rules. In subsequent years, more detailed modelling of the public sector would be explored.
- Documentation of programs
Public programs vary from country to country. It is essential that key features be described in the website.
Suggestion: UC-Berkeley come up with a template that would be part of the first year report.
- Tax incidence
We need to establish a "policy" about tax incidence. In general, we intend to follow the procedures in Generational Accounts. In most instances they assume that the tax is borne by the person who pays to the tax. Exception is taxes on capital in economies in which there are no barriers to international capital flows. Under these conditions, taxes on capital may be borne to some extent by workers in the form of lower wages. Hence, our policy about tax incidence may have to reflect the differences in the openess of economies.
Suggestion: Miller and others draft a policy; consult with Auerbach about final policy.
Private Transfers
- Time transfers
Time transfers are a very important component of familial transfers. The current system does not incorporate the value of time transfers. They should be considered in the future.
- Inter-household transfers
There is evidence that inter-household transfers are seriouly under-reported in many standard income and expenditure surveys. Detailed surveys of familial transfers conducted in some countries may provide a better source of information and allow some assessment of the extent to which traditional surveys provided under-estimates.
Suggestion: Perhaps this is an issue that could be explored by country teams with specialized surveys on familial transfers.
- Intra-household transfers
Methodology for estimating sector-specific intra-household transfers needs to be finalized. A promising suggestion is that within the household the same tax/subsidy system finance all transfers.
Suggestion: Mason will revise methodology in paper and test working with Nicole and Maliki.
- Capital transfers
A conceptual framework for modeling capital transfers was discussed at the meeting. This needs to be implemented and tested.
Suggestion: EWC team will work on the details of this.
Workplan/First Year Goals
Two goals for the first year were discussed at the meeting: (1) estimating the National Transfer Flow Account for at least one recent year; (2) projecting National Public Transfer Flow Accounts. Papers and presentations would be prepared for the 3d Workshop meeting. Following Generational Accounting Around the World, a common set of tables would be prepared by each team, and reports would also have a section for interesting features of the economy. Data would all be uploaded to the website.
Suggestion: EWC and UCB will draft a more detailed description of the material to be prepared during the first year.
Website
- Add search routines
- Add forum facility
- Encourage use by sending out frequent messages to project members
- Clarify details of the back-up
- Add information about procedures. A "walk through" which gives the step-by-step instructions for assemblying parts of the NTAs. Ron-- was this what you had in mind? - tim This sounds like including a tutorial or something like that. A good idea. Perhaps we could use some of the powerpoint slides from summer seminar or other workshops. andy
Rules of the Game
- How to cite
- When is permission needed
MEETING DOCUMENTS
Cover letter regarding the meeting
List of participant
Meeting's agenda
Presentations (Power point)
Andrew Mason, Overview of NTA, January 2005
Andrew Mason, Sang-Hyop Lee and Maliki, Consumption and Production, January 2005.
Andrew Mason with assistance from Mun-Sim Lai, Private Transfers Presentation, January 2005.
Ron Lee, Opening Remarks, Second Workshop on National Transfer Accounts, January 2005.
Cassio M. Turra, Reflections, A View from the South, January 2005.
Jorge Bravo, Data on Chile - an update, January 2005.
Cassio M. Turra and Bernardo L. Queiroz, National Transfer Accounts in Brazil, January 2005.
Rikiya Matsukura, National Transfer Accounts in Japan, January 2005.
Papers (Doc file)
Andrew Mason. An Overview of National Transfers Accounts, December 2004. NTA2004
Andrew Mason with assistance of Mun-Sim Lai, Measuring Private Transfers in National Transfer Accounts, January 2005.
Sang-Hyop Lee and Maliki, Cost of Children, January 2005.
An-Chi Tung, Some Notes on Building the National Transfer Accounts, January 2005. Tung2004
What would you think about simplifying the summary flow account by merging combining capital reallocations with property and credit reallocations - we could call these asset reallocations. There would be a detailed asset reallocation account that we could work on later. I think asset reallocations are pretty straight-forward to estimate directly. From the household's budget identity we know that:
rA - I = C - yl - transfers.
rA is non-labor factor income and I is investment (saving). We have macro controls for all of these. From the household surveys we have property income that is not retained but we could assume - as a first approximation - that the age distribution of distributed and retained earnings were identical. Thus, we could calculate rA. Then, I is estimated as a residual. If all of the elements in the equation have been adjusted to match their aggregate controls, I should be equal to aggregate saving.
This seems like a reasonable approximation. It allows us to complete an important part of the table while freeing us to focus our energy on public and private transfers. Andy
As I recall, the discussion at the meeting pretty much brought us to this conclusion. It is a simplification. Is anything important lost by this simplification? As I think about it, it becomes more difficult to reconcile the flow accounts with the estimates of the physical capital stock. In principle, if assets are estimated from the flow accounts, all the credit components should cancel whether or not they are separately identified, so assets should equal physical capital. But if credit inflows are under-reported less than outflows, or the reverse, then they won't aggregate up to 0 credit wealth, not to mention the issues of government debt, foreign holding of debt, etc. So there might be some loss from our inability to enforce the constraint that the total of credit derived from flows should equal govt debt less foreign holdings of govt debt, or something of that sort. Also, we would not be able to apply different rates of return to, say, govt bonds than to equity.
Also, we would not be able to say anything about the level of indebtedness by age, because all we would know would be something like net worth at each age.
But does any of this matter much? PRobably not. Does the gain in simplicity and in ease of explanation out weigh the loss in detail? Probably so.
I not completely comfortable with doing this, because I have not thought through all ramifications, but it did seem to be the sense of the meeting. Perhaps we could have a little more discussion among ourselves before making this definite.
Ron