Based on an analysis of recent scientific research, the Party has decided that the best contribution it can give to Israel is a proven technology to create peace in the Middle East, reduce crime and accidents, improve public health, and to make the nation self-sufficient. This technology is based on Maharishi's Vedic Science which includes among other programmes the Transcendental Meditation and TM-Sidhi.

More than 500 scientific research studies have been conducted at over 200 universities and research institutions in 30 countries, validating the benefits for mind, body, behaviour, and society, of the Transcendental Meditation® programme, as taught by Maharishi Mahesh Yogi. Within this large body of research, over 40 research studies during the past 20 years document highly significant positive effects for the whole society of individuals practising the Transcendental Meditation (TM®) technique and groups of individuals collectively practising the TM-Sidhi® programme. This influence of coherence and orderliness in society is known in the scientific literature as the Maharishi Effect.

The Maharishi Effect has been measured in terms of reduced crime, reduced traffic accidents and fatalities, reduced domestic violence and war conflict, improved quality of life, and improved economic trends. This effect has been documented at the city, state or provincial, national, and international levels during periods of time when large numbers of individuals gathered to collectively practising the Transcendental Meditation and TM-Sidhi programme. The scientific studies on the Maharishi Effect have been published in research journals such as Journal of Conflict Resolution, Social Indicators Research, The Journal of Mind and Behavior, and Proceedings of the American Statistical Association: Business and Economic Statistics Section.

The proportion of the population necessary to generate the Maharishi Effect has been found to be one per cent of the population practising the Transcendental Meditation programme individually in their own homes, or the square root of one per cent of the population practising the TM-Sidhi programme, which includes Yogic Flying, together in one group twice daily. The fact that the Maharishi Effect can be generated by a very small proportion of the population practising the TM-Sidhi programme together in one place makes it possible to implement a practical programme to reduce negative trends in the nation and improve the quality of national life. For example, the square root of one per cent of the approximately 7 million population of Israel is about 270 people. However, since Israel is a focal point of diverse political and religious interests of people from many countries, a group of 7000 participants in the group practice of the TM-Sidhi programme (approximately the square root of one per cent of the world’s population) is necessary in Israel to create a substantial influence of orderliness and harmony in the Middle East region and in the world as a whole.

The purpose of this project is to create a permanent group of experts in the TM-Sidhi (Yogic Flying) programme, to reduce previously intractable national problems and thereby reduce the expenses of the national government. The project will be paid for by cost savings provided to the Government of Israel. The Natural Law Partyt will use proceeds from the project to construct facilities and other necessities for the permanent group of 7000 experts.

The Party proposes the following stages as the framework for establishing its relationship with the Government of Israel and reliably evaluating the effects of the 7000 For Israel Project.

1. The Party will establish several teams of researchers, preferably one team from each University in Israel. One team will be established by the Party from international scientists who have previously conducted research on the Maharishi Effect. Each team will independently perform its research to evaluate the effects of the Project in different areas (please refer to the list in page 9 below) and the resulting cost saving to the National Economy, according to common guidelines for research established prior to the beginning of the Project (please refer to ‘Stage 3’, below), and according to a specified timeline for submitting results.
2. The Partywill approach the Government in order to reach an agreement for the implementation of the Project and sign a contract which will address the following issues:
1. The Government will assign a Supervisory Council composed of distinguished scientists and experts. We suggest that the Supervisory Council be chaired by:
   * The President of the Supreme Court.
   * The President of the National Academy of Sciences.
   * The Government Statistician.

1. Supervision of the implementation of the Project (see following section), including the validation of the number of participants and the collection from the Government of the statistical data from which the success of the Project will be evaluated;
1. Supervision of the work of the independent evaluation teams;
1. Receiving reports from the separate evaluation teams at the end of each step of the Project, within a specified period of time, and preparation of a final conclusion at each step of the Project;
1. Presentation of conclusion findings to the Government at the end of each step of the Project: including magnitude of change in negative trends and the resulting cost savings to the national economy.

1. According to the summary conclusions of the Supervisory Council at the end of each step of the project the government will participate in the financing of the Project, with a sum equal to 30% of the cost savings created by the just-completed step of the Project. These funds will be used for further development of the Project in order to secure and perpetuate the positive effects of the project for the nation.

1. Only after reaching the agreement with the Government, the various aspects of the Project will begin, as follows:
1. The evaluation teams and the Supervisory Council will finalize the methodology for the evaluation research, and will begin the collection and analysis of baseline data to confirm the variables to be included in the analysis of the Project.
2. The Party will secure the financial and personnel resources necessary to begin the implementation of the Project (please refer to steps of implementation immediately below).
3. The Party will begin to implement the Project with the first step of creating the first group of 1000 experts (see the following section).
2. At the end of each step of implementing the Project;
1. The evaluation teams receive data provided by the Government agencies, analyze the data, and present findings within a specified period of time to the Supervisory Council.
2. The Supervisory Council reviews the reports of the evaluation teams, and within a specified period of time presents final conclusions to the Government.
3. The Government contributes its share in financing of the project (please refer to Stage 2.b. above).

The Party will implement a 7-step project to establish a permanent group of 7000 experts in the TM-Sidhi programme in Israel. Each step will last six months. The 7000 For Israel Project is designed so that each step of the programme will bring measurable cost savings to the government of Israel; a proportion of this cost savings will be shared by the Party at the end of each step of the project in order to secure the stability of the 7000 group in perpetuity.

1. The first step of the project will be to bring 1000 previously trained experts in the TM-Sidhi programme to Israel, to form a group to immediately create an influence of coherence in the nation through the Maharishi Effect. The purpose of bringing previously trained experts during the initial phase of the project is that complete training in the Transcendental Meditation and TM-Sidhi programme is 4 months, in addition to the time needed to recruit the participants. (Instruction in the Transcendental Meditation technique takes one week. Individuals must practise the Transcendental Meditation technique for at least two months before being instructed in the TM-Sidhi programme, which takes two months.) During the six months of Step 1, at least 1400 Israelis will be recruited to be members of the permanent group of 7000 experts in the Transcendental Meditation and TM-Sidhi programme. The cost savings to the Government of Israel will be assessed at the end of Step 1 and a proportion shared by the Party.
   
2. During the second step of the project, the group of 1000 visiting experts will continue their practice of the TM-Sidhi (Yogic Flying) programme. At the same time the first 1400 Israelis will be instructed in the TM-Sidhi programme. At the end of the second step of six months, a portion of the continuing cost benefits from the original group of 1000 experts, accrued during these six months, will be received by the Party. The group of 1000 visiting experts in the TM-Sidhi programme will leave Israel at the end of Step 2.
   
3. The first group of 1400 Israelis trained in the TM-Sidhi programme forms the first part of the permanent group of TM-Sidhi Yogic Flying experts. The second group of 1400 Israelis (recruited during step 2 or before) will be trained during Step 3. At the end of the third step, a proportion of the cost benefits from the changes in national trends due to the group of 1400 practising experts accrued during these six months will be received by the Party.
   
4. Steps 4-7. Continuing in this way, additional groups of 1400 Israeli experts in the Transcendental Meditation and TM-Sidhi programme, 5 groups of 1400 experts in all, will be added until the total of a group of 7000 experts is achieved. A proportion of the additional cost savings to the national government during each six-month period will be received by the Party to secure the permanence of the group of 7000 Israeli experts in the Transcendental Meditation and TM-Sidhi programme.

Table 1 lists the 7 steps of the project, the number of people trained during each period, and the size of the total group of trained participants in the TM-Sidhi Yogic Flying practice (including those who come from outside Israel).

Table 1

Based on previous research studies on the Maharishi Effect, the Natural Law Party confidence that the measurable effects of the 7000 For Israel Project will include the following, each of which should bring significant cost savings for the government of Israel:

1. Reduced crime
2. Reduced motor vehicle accidents
3. Reduced injuries due to motor vehicle accidents
4. Reduced motor vehicle fatalities
5. Reduced fires
6. Reduced hospital admissions
7. Reduced absenteeism from work
8. Reduced inflation
9. Reduced unemployment
10. Reduced worker days lost in strikes
11. Reduced injuries and fatalities due to terrorism or domestic violence

Any of these variables that meet the criteria of proper statistical suitability established by the Supervisory Council and evaluation teams will be included in the evaluation. The Appendix on the following pages gives a proposal for evaluation and statistical analysis that will serve as a starting point for discussion when the Supervisory Council and evaluation teams begin to determine the final protocol for evaluation.

It is proposed that the government of Israel and the Foundation contract for the implementation of this project. At the end of each of the 7 six-month steps of the project, the government of Israel will share with the Foundation the cost savings to the nation for each of the variables evaluated for which a statistically significant improvement has been found. For each of these variables the amount of the Foundation’s share is proposed to be 30% of the cost savings to the government resulting from the measured percentage improvement. The amount to be paid for a given percentage of improvement on a variable is based on cost-savings calculations made by independent researchers in advance of the project and agreed upon by the government, the Supervisory Council, and the Foundation.

It is proposed that the government of Israel and the Party contract for the implementation of this project. At the end of each of the 7 six-month steps of the project, the government of Israel will share with the Party the cost savings to the nation for each of the variables evaluated for which a statistically significant improvement has been found. It is proposed that the government of Israel and the Foundation contract for the implementation of this project. At the end of each of the 7 six-month steps of the project, the government of Israel will share with the Foundation the cost savings to the nation for each of the variables evaluated for which a statistically significant improvement has been found. For each of these variables the amount of the Party's share is proposed to be 30% of the cost savings to the government resulting from the measured percentage improvement. The amount to be paid for a given percentage of improvement on a variable is based on cost-savings calculations made by independent researchers in advance of the project and agreed upon by the government, the Supervisory Council, and the Party.

Based upon previous research, it is estimated that the magnitude of reduction of negative trends will be between 15% and 20% for the first group of 1000 participants in the project (year 1 - steps 1 and 2), rising to between 60% and 80% at the end of step 7 (year 3 - 7000 participants), with further improvement continuing as the group of 7000 is maintained.

This Appendix is included to provide a starting point for the determination of the final evaluation protocol by the evaluation teams and the Supervisory Council, and to give an idea of the type of rigorous evaluation expected by the Party.

As noted in the main text, previous research on the Maharishi Effect indicates that the following variables should be significantly influenced in the positive direction through the 7000 For Israel Project: (1) reduced crime; (2) reduced motor vehicle accidents; (3) reduced injuries due to motor vehicle accidents; (4) reduced motor vehicle fatalities; (5) reduced fires; (6) reduced hospital admissions; (7) reduced absenteeism from work; (8) reduced inflation; (9) reduced unemployment; (10) reduced worker days lost in strikes; (11) reduced injuries and fatalities due to terrorism or domestic violence; .

It is proposed that in the initial stage of the evaluation process, baseline data on each variable is collected and the behaviour of the data reviewed to ensure that the variable is suitable for statistical evaluation during the Project period. The criteria for statistical suitability of a variable may include the following: (1) the ability to find appropriate statistical models to describe the data prior to the Project period; (2) the ability to include appropriate independent or control variables that influence the variable prior to the Project period, if such independent variables exist; and (3) the stability of the time series of data immediately prior to the Project period. In the case that there is an instability of one or more of the variables prior to the Project period, it may be possible to analyze the effects on that variable using a shorter baseline of data that may still be of sufficient length to provide adequate statistical power.

It is proposed that the data from variables 1-6 be collected on a daily basis over a 6-year baseline period prior to the start of the project, and aggregated into weekly totals to give a series of over 300 data points. This is done because weekly aggregation will improve the stability of the data and remove the substantial weekly cycles in many of these variables. A series of this length should provide suitable statistical power. For the economic variables 7-9, data is usually available on a monthly basis; for this reason at least 20 years (240 data points) of monthly baseline data prior to the start of the project will be collected. For variables 10 and 11 (worker days lost in strikes and injuries and fatalities due to terrorism or domestic violence), there may be substantial discontinuities in the data, and thus monthly aggregation is also recommended to create a more stable series of data, with 20 years (240 data points) of monthly baseline data collected prior to the start of the project.

Several additional analyses may be performed not for the sake of determining the cost benefit impact of the Project but to determine broader dimensions of the Project's effects. For example, where possible, an overall index composed of as many of the previously-specified variables as possible may be constructed and evaluated statistically to determine that the transformation of national life is comprehensive and not limited to the isolated factors that might influence only single outcome variables. Similarly, if they desire, specific evaluation teams may wish to include an evaluation of more subjective aspects of national quality of life to the extent that such data is readily available.

The reduction in negative trends for each variable at each of the seven steps of the project will be assessed through time series analysis using the autoregressive integrated moving average (ARIMA) methodology (Box and Jenkins, 1976). This methodology is considered as the most rigorous for precisely assessing intervention influences on a time series or for empirically evaluating the form of causal relationship between two time series (McCleary and Hay, 1980).

Time series intervention analysis will be used to evaluate the effect of each separate step of the project on the dependent or exogenous social indicator variables. This analysis will define as an intervention the weeks in which the number of group participants is over a given threshold. In this case, the threshold of the square root of one per cent of the 7 million population of Israel - about 270 people - should be exceeded from the time the first group of 1000 TM-Sidhi participants is established, i.e., throughout Step 1 of the project. Thus, for the first step of the project, the period of establishment of the group will be defined as the intervention period in contrast to the baseline period prior to the establishment of the group. The effect of the intervention is estimated while modeling and thus controlling for any seasonal patterns of the endogenous variables. Each of the subsequent phases of the project, during which the size of the group is sequentially expanded, can also be evaluated as a separate intervention, and the magnitude of its effects separately calculated.

Time series transfer function analysis will also be used for secondary analyses at later stages of the project. In the transfer function analysis approach, the exogenous or independent variable is continuous (in this case the number of TM-Sidhi programme experts) and the analysis models the input-output relationship between the exogenous and the endogenous variable. In this way the change in the endogenous variable, based on the exogenous one, is assessed while controlling for the internal dynamics of each variable over time. The transfer function approach will be appropriate for assessing the effects of the increasing size of the group over time, at approximately six-month intervals.

With both approaches, the time series methodology controls for any serial dependence of observations, long-term nonstationarity or trends, or seasonal cycles in the data over time, by including these influences in a "noise model" of the series (McCleary and Hay, 1980) which serves as the null hypothesis for effects of the exogenous variable. Observed intervention effects or transfer function effects on the endogenous variable would indicate effects of the exogenous variable that cannot be predicted either from the previous history of the series or from any unmeasured continuous variables that may be partially determining the endogenous variable. The noise model Nt has a form Nt=[f (B)]-1q (B)at, where f (B) and q (B) specify autoregressive and moving average parameters, respectively, at various time lags, B is a backshift operator, and at is a series of independent and normally distributed random disturbances. The noise model effectively removes the serial dependence of the data by modeling it, and the residuals to the noise model (at) form independent data points for which parametric statistical models are appropriate.

In transfer function analysis, the endogenous time series Yt is modeled as Yt=C+V(B)Xt+Nt, where Xt is the continuous exogenous series, V(B) is a transfer function connecting the two series, C is a constant, if necessary, and Nt is a stochastic noise model, defined above, specifying the combined nonrandom influences other than the exogenous series (Box and Jenkins, 1976). The intervention analysis model is identical, except that the exogenous variable is the binary intervention series It. The exogenous effect V(B) is comprised of impulse response weights vi, such that V(B)=v0+v1B+v2B2+.... This function is approximated by a rational polynomial of the form [d (B)]-1w (B), where w (B) contains parameters indicating the time delay of influence of the exogenous variable and the magnitude of its effect at various time lags, and where d (B) contains parameters specifying the rate at which this influence decays (for an abrupt temporary effect) or grows (for a gradual permanent effect) (Box and Jenkins, 1976).

Transfer function models may be identified and estimated using the linear transfer function (LTF) approach of Liu and Hanssens (1982). The LTF method directly estimates the impulse response weights as distributed lagged effects of the exogenous series (Liu and Hanssens, 1982). Identification of the transfer function is determined by estimating the equation Yt=C+V(B)It+Nt or Yt=C+V(B)Xt+Nt to obtain the impulse response weights V(B). In the LTF approach, the number of lags over which the impulse response weights are estimated is chosen on the basis of subject-matter considerations, and should be sufficient to avoid truncation bias.

In this project, effects are predicted relatively close in time (an almost immediate effect on the behavioural variables), and lagged effects up to six weeks (lags 0 to 6) will be examined for weekly data. (For the variables of unemployment, inflation, and worker days lost in strikes, which will be available on a monthly basis and for which previous research has shown longer lagged effects, lagged effects up to six months (lags 0 to 6) will be examined; this will mean that some of the full effects on these monthly variables may only be determined one year after a given intervention.

After identifying the transfer function for It or Xt, the appropriate model parameters will be estimated, and residuals to the transfer function will be used to identify the noise model further in the iterative manner described by Box and Jenkins (1976) and McCleary and Hay (1980). Final parameter estimates will use the maximum likelihood method. Moving average parameters will be estimated by an 'exact' likelihood function of Hillmer and Tiao (1979) described and implemented in Liu and Hudak (1986). A diagnostic test of the joint significance of residual autocorrelations is given by Ljung and Box (1978).

In the case that there are multiple significant intervention parameters (e.g., effects of a single intervention at various time lags, or effects of the series of interventions - steps 1-7), the combined statistical significance of all intervention parameters will be computed from a likelihood ratio test given by Nelson (1976).

An objective criterion of model appropriateness, the Akaike 'information criterion' (AIC) (Akaike, 1973; Larimore and Mehra, 1985), can be used at a final stage of model selection. The statistical foundations of the AIC are developed in Larimore (1983). To increase the objectivity of the model selection process, the AIC will be used only to select a final model from among several models that are all acceptable according to other conventional diagnostic tests, and thus to avoid the possibility of biasing the results of the main analysis due to arbitrary selection of a noise model. It is anticipated, however, that the results will be robust across all plausible alternative noise models, and therefore will not depend on the specific model selection procedure.

When models are compared using the AIC criterion, the models will be estimated on the same sample of effective observations, since the AIC is dependent upon sample size and estimation of autoregressive parameters consume larger numbers of observations in model estimation than moving average parameters.

Because the steps of implementation of the project will occur every six months, to control conservatively for long-term seasonality in the series of weekly data while assessing the effect of the six-month interventions, the autocorrelations will be modeled up to lag 60 in the evaluation of noise models.

Reported p values for parameter estimates will be based upon two-tailed tests for all noise model parameters and constants, and one-tailed tests for intervention parameters, since the direction of effect is clearly predicted. Statistical significance is defined as p < .05.

Akaike, H. (1973). Information theory and an extension of the maximum likelihood principle. In B. A. Petrov & F. Csaki. (eds.), International symposium on information theory (pp. 267-281). Budapest: Akademia Kiado.

Box, G. E. P., & Jenkins, G. M. (1976). Time series analysis: forecasting and control. San Francisco: Holden-Day.

Chalmers, R. A., Clements, G., Schenkluhn, H., & Weinless, M. (eds.). (1989). Scientific research on Maharishi's Transcendental Meditation and TM-Sidhi program: Collected papers, Volumes 2–4. Vlodrop, the Netherlands: Maharishi Vedic University Press.

Hillmer, S. C. & Tiao, G. C. (1979). Likelihood function of stationary multiple autoregressive moving average models. Journal of the American Statistical Association, 74, 652-660.

Larimore, W. E. (1983). Predictive inference, sufficiency, entropy, and an asymptotic likelihood principle. Biometrika, 70, 175-181.

Larimore, W. E. & Mehra, R. K. (1985). The problem of overfitting data. Byte, 10(10), 167-180.

Liu, L.-M., & Hanssens, D. M. 1982. Identification of multiple-input transfer function models. Communications in Statistics A, 11, 297-314.

Liu, L.-M. & Hudak, G. B. (1986). The SCA Statistical System: Reference manual for forecasting and time series analysis. DeKalb, IL: Scientific Computing Associates.

Ljung, G. M., & Box, G. E. P. (1978). On a measure of lack of fit in time series models. Biometrika, 65, 297-303.

Maharishi Effect. (1990). Fairfield, Iowa: Maharishi International University Press.

McCleary, R., & Hay, R. A. Jr. (1980). Applied time series analysis for the social sciences. Beverly Hills, CA: Sage.

Nelson, C. R. (1976). The interpretation of R2 in autoregressive-moving average time series models. The American Statistician, 30, 175-180.

Orme-Johnson, D. W., & Farrow, J. T. (eds.) (1976). Scientific research on the Transcendental Meditation program: Collected papers, Volume 1. Rheinweiler, W. Germany: Maharishi European Research University Press.

Scientific research on Maharishi's Transcendental Meditation and TM-Sidhi program: Collected papers, Volume 6. (in press). Vlodrop, the Netherlands: Maharishi Vedic University Press.

Wallace, R. K., Orme-Johnson, D. W., & Dillbeck, M. C. (eds.). (1990). Scientific research on Maharishi's Transcendental Meditation and TM-Sidhi program: Collected papers, Volume 5. Fairfield, IA: Maharishi International University Press.